Remarks on the Gauss-Green Theorem. Michael Taylor
|
|
- Jasper Charles
- 6 years ago
- Views:
Transcription
1 Remarks on the Gauss-Green Theorem Michael Taylor Abstract. These notes cover material related to the Gauss-Green theorem that was developed for work with S. Hofmann and M. Mitrea, which appeared in [HMT]. Contents 1. Versions of the Gauss-Green formula 2. Examples of domains with locally finite perimeter 3. Green formulas for layer potentials 4. Layer potentials on Ahlfors regular sets; elementary estimates 5. Green s formula on Ahlfors regular domains 6. Gauss-Green formulas on Riemannian manifolds A. The Gauss-Green formula on smoothly bounded domains B. Variant of a result of Chen and Torres 1. Versions of the Gauss-Green formula Let R m be open. We say has locally finite perimeter provided (1.1) χ = µ is a locally finite R m -valued measure. It follows from the Radon-Nikodym theorem that µ = ν σ, where σ is a locally finite positive measure, supported on, and ν L (, σ) is an R m -valued function, satisfying ν(x) = 1, σ-a.e. It then follows from the Besicovitch differentiation theorem that 1 (1.2) lim ν dσ = ν(x) r 0 σ(b r (x)) B r (x) for σ-a.e. x. Via distribution theory, we can restate (1.1) as follows. Take v C 0 (R m, R m ) (a vector field). Then (1.3) div v, χ = v, χ. Hence (1.1) is equivalent to (1.4) div v dx = ν v dσ. 1
2 2 Works of Federer and of De Giorgi produced the following results on the structure of σ, when has locally finite perimeter. First, (1.5) σ = H m 1, where H m 1 is (m 1)-dimensional Hausdorff measure and is the reduced boundary of, defined as the set of points x such that the limit (1.2) holds, with ν(x) = 1. (It follows from the remarks leading up to (1.2) that σ is supported on.) Second, is countably rectifiable; it is a countable disjoint union (1.6) = k M k N, where each M k is a compact subset of an (m 1)-dimensional C 1 surface (to which ν is normal in the usual sense), and H m 1 (N) = 0. Given (1.5), the identity (1.4) yields the Gauss-Green formula (1.7) div v dx = ν v dh m 1, for v C 0 (R m, R m ). It is also useful to record some results on sets 0. Good references for this material, as well as the results stated above, are [EG], [Fed3], and [Zie]. First, given a unit vector ν E and x, set (1.8) H ± ν E (x) = {y R m : ±ν E (y x) 0}. Then (cf. [EG], p. 203), for x, + =, = R m \, one has (1.9) lim r 0 r m L m( B r (x) ± H ± ν E (x) ) = 0, when ν E = ν(x), as in (1.2). Here L m denotes Lebesgue measure on R m. More generally, a unit vector ν E for which (1.9) holds is called the measure-theoretic outer normal to at x. It is easy to show that if such ν E exists it is unique. Thus if we define 0 to consist of x for which (1.9) holds, with ν E (x) denoting the measure-theoretic outer normal, we have 0 and ν E (x) = ν(x) on. Next, we define, the measure-theoretic boundary of, to consist of x such that (1.10) lim sup r 0 r m L m( B r (x) ±) > 0. It is clear that 0. Furthermore (cf. [EG], p. 208) one has (1.11) H m 1 ( \ ) = 0.
3 3 Consequently the Green formula (1.7) can be rewritten (1.12) div v dx = ν v dh m 1, for v C 0 (R m, R m ). The advantage of (1.12) is that the definition of is more straightforward and geometrical than is that of. Remark. Note that is well defined whether or not has locally finite perimeter. It is known that has locally finite perimeter if and only if H m 1 ( K) < for each compact K R m. (Cf. [EG], p. 222.) In general \ can be quite large. It is of interest to know conditions under which H m 1 ( \ ) = 0. One such result will be given in 2. So far we have discussed the Green formula for v C 0 (R m, R m ). A simple limiting argument extends (1.4) and hence (1.7) to v C 1 0(R m, R m ); [Fed] emphasizes that (1.7) is true for compactly supported Lipschitz v. In fact, we can easily do better. Here is one improvement. Proposition 1.1. If has locally finite perimeter, then (1.4) holds for v in (1.13) D = {v C 0 0(R m, R m ) : div v L 1 (R m )}. Proof. Given v D, take ϕ C 0 (R m ) such that ϕ dx = 1, set ϕ k (x) = k m ϕ(kx), and take v k = ϕ k v C 0 (R m ). Then (1.4) applies to v k, i.e., (1.14) div v k dx = ν v k dσ. Meanwhile, div v k = ϕ k (div v) div v in L 1 (R m ) and ν v k ν v uniformly on, so as k, the left side of (1.14) converges to the left side of (1.4) while the right side of (1.14) converges to the right side of (1.4). In many cases one deals with functions defined only on, and one would like to avoid assuming they have extensions to R m with nice properties. To obtain a result for such functions, we will introduce the following concept. Let open sets k satisfy k, k k+1, and k. We say { k : k 1} is a tame interior approximation to if in addition there exists C(R) < such that, for R (0, ), (1.15) χ k TV(BR ) C(R), k 1. To give an example, take A : R m 1 R, satisfying (1.16) A C(R m 1 ), A L 1 loc(r m 1 ).
4 4 As will be shown in 2, (1.17) = {(x, x m ) R m : x m > A(x )} has locally finite perimeter. The estimates proven there then imply that (1.18) k = {(x, x m ) R m : x m > A(x ) + k 1 } is a tame interior approximation to. Proposition 1.1. The following is a partial extension of Proposition 1.2. Assume has locally finite perimeter and a tame interior approximation. Then (1.4) holds for v in (1.19) D = {v C 0 0 (, R m ) : div v L 1 ()}. Proof. Let { k } denote a tame interior approximation. Pick ϕ k C0 () to be 1 on a neighborhood of k supp v, set v k = ϕ k v, and apply Proposition 1.1 with replaced by k and v by v k, noting that div v k = ϕ k div v + ϕ k v. We have (1.20) div v dx = v, χ k. k As k, the left side of (1.20) converges to the left side of (1.4). Meanwhile, we can take w C 0 0(R m, R m ), equal to v on, and the right side of (1.15) is equal to w, χ k. Now χ k χ in L 1 loc (Rm ), so χ k χ in D (R m ), and hence (1.21) w, χ k w, χ for each w C 0 (R m, R m ). The bounds (1.15) then imply that (1.21) holds for each w C 0 0(R m, R m ). Hence the right side of (1.20) converges to (1.22) w, χ = which is the right side of (1.4). ν v dσ, As we will discuss in 3, Proposition 1.2 is not adequate for the Green formulas we need for layer potentials. Such a Green formula will be demonstrated in 5. Remark. Proposition 1.2 can be compared with the following result, given in
5 [Fed1], p Let R m be a bounded open set such that H m 1 ( ) <. Fix j {1,..., m} and take f such that (1.23) f C(), j f L 1 (). 5 Then (1.24) j f dx = e j ν f dh m 1, 0 where e j is the jth standard basis vector of R m. In light of (1.11) one could replace 0 by or by in (1.24). This leads to the identity (1.7) for a vector field v C() provided each term j v j in div v belongs to L 1 (). However, the vector fields arising in the applications of Green s formula needed in 3 and 5 need not have this additional structure, so (1.24) is not applicable. There is also recent work of [CT] and [CTZ], which we will briefly discuss in Appendix B. We turn our attention to a Green formula for (1.25) div v dx, B r where has locally finite perimeter and B r = {x R m : x < r}. Assume v C 0,1 0 (Rm, R m ). Given ε (0, r), set ψ ε (x) = 1 for x r ε, (1.26) 1 1 ( x r + ε) for r ε x r, ε 0 for x r. Then, with χ = ν σ, we have div v dx = lim (1.27) B r ε 0 ε 0 ψ ε div v dx = lim [div ψ ε v v ψ ε ] dx ( = lim ε 0 = B r ν ψ ε v dσ 1 ν v dσ + lim ε 0 ε S ε ) v ψ ε dx n v dx,
6 6 where n is the outward unit normal to B r and (1.28) S ε = B r \ B r ε. Consequently, (1.29) div v dx = B r B r ν v dσ + D r Φ(r), where (1.30) Φ(r) = B r n v dx. Note that (1.31) Φ(r) = r 0 B s n v dh m 1 ds, so (1.32) D r Φ(r) = B r n v dh m 1, for L 1 -a.e. r. It is of interest to note that D r Φ(r) exists for all r (0, ) (under our standing hypothesis on ), though the identity (1.32) is valid perhaps not for each r, but just for a.e. r. Remark. Having (1.29), one can bring in (1.5) and write (1.33) B r div v dx = B r ν v dh m 1 + D r Φ(r). It is important to note that (1.5) is not needed to prove (1.29), since (1.29) plays a role in proofs of (1.5).
7 7 2. Examples of domains with locally finite perimeter Let be the region over the graph of a function A : R m 1 R: (2.1) = {x R m : x m > A(x )}, where x = (x, x m ). We have: Proposition 2.1. If (2.2) A C(R m 1 ), A L 1 loc(r m 1 ), then has locally finite perimeter. Proof. Pick ψ C 0 (R m 1 ) such that ψ(x ) dx = 1, set ψ k (x ) = k m 1 ψ(kx ), and set A k = ψ k A and (2.3) k = {x R m : x m > A k (x )}. Clearly (2.4) A k A, locally uniformly, so (2.5) χ k χ in L 1 loc(r m ). Hence χ k χ in D (R m ). Also (2.6) χ k = ν k σ k, where σ k is surface area on (2.7) Σ k = {x R m : x m = A k (x )}, given in x -coordinates by (2.8) dσ k (x ) = 1 + A k (x ) 2 dx, and ν k is the downward-pointing unit normal to Σ k. The hypothesis (2.2) implies that {ν k σ k : k 1} is a bounded set of R m -valued measures on each set B R = {x R m : x R}, so passing to the limit gives (2.9) χ = µ
8 8 where µ is a locally finite R m -valued measure. This proves Proposition 2.1. Remark. The proof above is an easy extension of the argument in [Zie] that the domain over a Lipschitz graph has locally finite perimeter. The measure µ in (2.9) has the form µ = ν σ, as described after (1.1). To obtain a more explicit formula, we invoke (1.4), (2.10) div v dx = ν v dσ, together with the elementary identities (2.11) div v dx = ( A k (x ), 1) v(x, A k (x )) dx, k R m 1 valid for each v C0 (R m, R m ). (See Appendix A for an elementary proof.) As k, the left side of (2.11) converges to the left side of (2.10), while the right side of (2.11) converges to (2.12) ( A(x ), 1) v(x, A(x )) dx. R m 1 Hence (2.13) where ν v dσ = R m 1 ν(x ) v(x, A(x )) dσ(x ), (2.14) ν(x ) = ( A(x ), 1) 1 + A(x ) 2, dσ(x ) = 1 + A(x ) 2 dx. The formula (2.13) is valid for all v C0 (R m, R m ), hence for all v C0(R 0 m, R m ). Remark. So far in this section we have not used (1.5). At this point we can invoke (1.5), to get (2.15) ν v dh m 1 = ν(x ) v(x, A(x )) dσ(x ), R m 1 for each v C0(R 0 m, R m ).
9 It is also of interest to see how the decomposition (1.6), asserting countable rectifiability, arises in the context of (2.1) (2.1). For simplicity, assume A has compact support. Then set (2.16) f = A + A, g = Mf, the latter being the Hardy-Littlewood maximal function, and take (2.17) R λ = {x R m 1 : Mf(x) λ}. Then L m 1 (R m 1 \ R λ ) Cλ 1 f L 1, and an argument involving the Poincaré inequality yields (2.18) x, y R λ = A(x), A(x) A(y) Cλ x y. Using this one writes = k K k Ñ, where each L k is a Lipschitz graph and σ(ñ) = 0. Passing to (1.6) is then done by decomposing each Lipschitz graph into a countable union of C 1 graphs plus a negligible remainder, via Rademacher s theorem and Whitney s extension theorem. See 6.6 of [EG], or Theorem 11.9 and Proposition 12.9 of [T], for details. Regarding the issue of versus, it is clear that = whenever A is locally Lipschitz. For more general A satisfying (2.2), we have the following, which is a consequence of the main results of [Tom] and [Fed2]. Proposition 2.2. If is the region in R m over the graph of a function A satisfying (2.2), then (2.19) H m 1 ( \ ) = 0. Proof. Given a rectangle Q = I 1 I m 1 R m 1, a product of compact intervals, set K Q = Q R. Given the formula (2.14) for σ, it follows from Theorem 3.17 of [Tom] that (2.20) σ( K Q ) = I m 1 ( K Q ), where I m 1 denotes (m 1)-dimensional integral-geometric measure. Furthermore, it is shown in [Fed2] that (2.21) H m 1 ( K Q ) = I m 1 ( K Q ). On the other hand, we have from (1.5) that σ( K Q ) = H m 1 ( K Q ), so (2.19) follows. 9 Remark. The main argument in [Fed2] showed that there exists C m < such that (2.22) H m 1 (S) C m I m 1 (S),
10 10 for all Borel sets S K Q. Federer then invoked his structure theorem to deduce that K Q is countably rectifiable and (2.21) holds. We can avoid depending on the structure theorem as follows. Theorem 3.17 of [Tom] also says that if we have a Borel set S K Q and set S b = {x Q : (x, A(x)) S}, then, under the hypothesis (2.2), (2.23) L m 1 (S b ) = 0 = I m 1 (S) = 0. Now as discussed above, ( \ ) Q A(V ) = {(x, A(x)) : x V } with V of Lebesgue measure 0, so (2.22) (2.23) give (2.19).
11 11 3. Green formulas for layer potentials Assume R m is a bounded NTA domain, and satisfies the Ahlfors regularity condition: there exist C j (0, ) such that (3.1) C 1 r m 1 H m 1 ( B r (p)) C 2 r m 1, for each p, r (0, diam ]. More generally, can be a UR domain, as defined in 3 of [HMT]. Take f L 2 (, σ), with σ = H m 1, and form (3.2) u(x) = Sf(x), where S is the single layer potential. Then, for a.e. x, Sf(x) = Sf(x) and ( (3.3) lim ν(x) u(y) = T f(x) = 1 y x 2 + K ) f(x), the limit being from within via a nontangential approach. We would like to be able to establish that (3.4) u 2 dx = u T f dσ. Under these hypotheses, = (modulo a null set), and if we set (3.5) v(x) = u(x) u(x), x, we have (3.6) div v = u 2 + u u = u 2 on, and, for x, (3.7) ν(x) lim y x in Γ(x) v(y) = u(x)t f(x). Hence (3.4) would follow from (1.4), if (1.4) were established in this context. For such, it is a fundamental fact, proven in [HMT], that (3.8) N ( Sf) L p ( ) C p f L p ( ), 1 < p <. A more elementary estimate is (3.9) N Sf(x) C x y (m 2) f(y) dσ(y),
12 12 if m 3 (with a simple modification for n = 2), which in turn yields, for all δ > 0, (3.10) N Sf L (m 1)/(m 2) ( ) C δ f L 1+δ ( ), N Sf L ( ) C δ f L m 1+δ ( ), when (3.1) holds. Interpolation gives (3.11) N Sf L 2+α(m) ( ) C f L 2 ( ), with α(m) > 0. In 4 we will establish a result that implies (3.12) Sf L p () C f L 2 ( ), p < 2m m 1. A similar analysis yields an estimate (3.13) Sf L q () C f L 2 ( ), for somewhat larger q. Thus, if v is given by (3.5), with u = Sf, we have (3.14) N v L p ( ) C f 2 L 2 ( ), p = p m > 1, and (3.15) div v L q () C f 2 L 2 ( ), q = q m > 1. Thus to establish (3.4) it will suffice to prove that (3.16) div v dx = ν v dσ, provided that, for some p > 1, (3.17) v L p, and div v L 1 (), where (3.18) L p = {v C() : N v L p ( ), and nontangential limit v b, σ-a.e.}. This will be proven for satisfying (3.1) in 5.
13 13 4. Layer potentials on Ahlfors regular sets: elementary estimates Let be a bounded open set. Assume satisfies the Ahlfors regularity condition (4.1) C 1 r m 1 H m 1 ( B r (p)) C 2 r m 1, p. We estimate some integral operators mapping functions on to functions on. Let (4.2) K : R be continuous on (4.3) \ {(x, x) : x } and satisfy (4.4) K(x, y) C x y (m 1). Define (4.5) T f(x) = where dσ = dh m 1. We aim to prove: K(x, y)f(y) dσ(y), Proposition 4.1. Under the hypotheses on and K given above, we have for p [1, ), (4.6) T : L p ( ) L r (), r < pm m 1. Proof. It is elementary that (4.7) T : L 1 ( ) L q (), q < m m 1. For this one needs only bounded and H m 1 ( ) <. We will show that (4.8) T : L ( ) L s (), s <. Then (4.6) follows by interpolation from (4.7) (4.8). To prove (4.8), we will establish an estimate of the form (4.9) K(x, y) dσ(y) γ(x), γ L s (), s <. Then (4.10) and (4.8) follows. Here is part of (4.9). K(x, y)f(y) dσ(y) f L γ(x),
14 14 Proposition 4.2. In the setting of Proposition 4.1, (4.11) K(x, y) dσ(y) C log 2M ϕ(x), ϕ(x) = dist(x, ), with M = diam. Proof. Given x, δ = dist (x, ), take p, x p = δ, and set (4.12) A 0 = {x : x p 2δ}, A k = {x : x p (2 k δ, 2 k+1 δ]}, k 1. By (4.1), (4.13) H m 1 (A k ) C(2 k δ) m 1, so (4.14) A k (2 k δ) m 1 K(x, y) dσ(y) C (2 k = C. δ) m 1 Summing (4.14) over k 0 such that 2 k δ M gives (4.11). The other half of (4.9), i.e., the fact that γ L s () for all s <, follows from: Lemma 4.3. Let R m be a bounded open set whose boundary is Ahlfors regular. Set (4.15) O δ = {x : dist(x, ) δ}. Then (4.16) vol(o δ ) Cδ. Proof. Let B 1 be the unit ball in R m, with volume C m, and set (4.17) χ δ (x) = 1 ( x ) C m δ m χ B 1, δ so χ δ dx = 1. Set µ = H m 1 and (4.18) G δ = µ χ δ. Then (4.19) G δ (x) dx = H m 1 ( ), δ > 0.
15 15 We also have the following. If p, x p = dist(x, ), (4.20) dist (x, ) δ 2 G δ(x) Cδ m H m 1 ( B δ/2 (p)) Cδ 1, the last inequality by (4.1). Hence (4.21) vol(o δ/2 ) Cδ G δ (x) dx Cδ H m 1 ( ). O δ/2 This proves (4.16). Thus (4.9) is proven, and hence so is Proposition 4.1. It is of incidental interest that Lemma 4.3 has the following corollary, related to the setting of Proposition 1.2. Corollary 4.4. If R m is a bounded open set and satisfies (4.1), then has a tame interior approximation. Proof. Consider ϕ Lip() given by ϕ(x) = dist(x, ), and set s = {x : ϕ(x) s}. For δ > 0, set (4.22) ψ δ (x) = δ on δ ϕ(x) on \ δ. Thus ψ δ is supported on O δ and ψ δ = 1 on O δ. A version of the co-area formula (Theorem of [Zie]) gives (4.23) ψ δ dx = O δ δ 0 χ s T V ds. Thus, by (4.16), since the left side of (4.23) equals vol (O δ ), (4.24) δ 0 χ s T V ds Cδ. Thus, for each k 1, there exists s (0, 1/k) such that χ s T V proves the corollary. C. This
16 16 5. Green s formula on Ahlfors regular domains Let R m be a bounded open set with finite perimeter. Assume (5.1) H m 1 ( \ ) = 0, and denote H m 1 by σ. For p [1, ), set (5.2) L p = {v C() : N v L p ( ), and nontangential limit v b, σ-a.e.}. Here L p ( ) = L p (, dσ). We will establish a Green formula for vector fields v L p with divergence in L 1 (), when satisfies the following two conditions. First, (5.3) 1 v dx C N v ( ), δ v L 1, 0 < δ diam, O δ where O δ = {x : dist(x, ) δ}. Second, (5.4) If v L p, w L 1 such that w b = v b and w k Lip() such that N (w w k ) L 1 ( ) 0. Then we will show that (5.3) and (5.4) hold whenever is Ahlfors regular. Before we state our first result, a few comments are useful. First, the condition (5.3) is equivalent to the apparently stronger condition (5.5) where 1 v dx C N δ v ( ), δ v L 1, 0 < δ diam, O δ (5.6) N δ v(x) = sup { v(y) : y Γ x, x y 2δ}, as a simple cutoff argument shows. Second, in condition (5.4) it would be equivalent to demand merely that w k C(), since elements of C() are easily uniformly approximated by Lipschitz functions. Here is our first result.
17 Proposition 5.1. Pick p [1, ). Assume is a bounded open set with finite perimeter, satisfying (5.1) and (5.3) (5.4). Then (5.7) div v dx = ν v b dσ, 17 whenever (5.8) v L p and div v L 1 (). Proof. Let s = {x : dist(x, ) s}. Let ϕ δ (x) = dist(x, δ/2 ), and set (5.9) χ δ (x) = 1 on δ, 2δ 1 ϕ δ (x) on \ ( δ O δ/2 ), 0 on O δ/2. If v satisfies (5.8), then χ δ v C0() 0 and div χ δ v L 1 (), so it is elementary that (5.10) div(χ δ v) dx = 0. Hence χ δ div v dx = χ δ v dx (5.11) = 2 δ ν v dx, Õ δ where ν = ϕ 0 and Õδ = O δ \ O δ/2. The left side of (5.11) converges to the left side of (5.7) as δ 0, whenever div v L 1 (). Hence (5.7) is true provided 2 (5.12) ν v dx ν v b dσ, δ Õ δ as δ 0. Of course, by (5.11), the left side of (5.12) does converge as δ 0, namely to the left side of (5.7). Hence (5.12) is true whenever (5.7) is true. In particular, (5.12) is true whenever v Lip(). More generally, if v satisfies (5.8), take w, w k as in (5.4). Given (5.3), we have (5.13) 2 δ Õ δ (ν w k ν w) dx C N (w w k ) L 1 ( ),
18 18 and we also have (5.14) (ν w k ν w b ) dσ N (w k w) L 1 ( ). Thus, since (5.12) holds for w k, we have 2 ν w dx δ (5.15) Õ δ = ν w b dσ ν v b dσ, as δ 0. Thus to obtain (5.12) for each v L p, it suffices to show that 2 (5.16) δ Õ δ (ν v ν w) dx 0. Thus it suffices to show that (5.17) u L 1, u b = 0 = 2 u dx 0, δ Õ δ as δ 0. Recalling that (5.3) implies (5.5), we see that it suffices to show that (5.18) u L 1, u b = 0 = N δ u L 1 ( ) 0, as δ 0. Indeed, the hypotheses of (5.18) yield N δ u(x) 0, σ-a.e., and furthermore N δ u N u for each δ, so (5.18) follows from the dominated convergence theorem. Proposition 5.1 is proven. We next show that Ahlfors regularity implies (5.4). Proposition 5.2. If R m is a bounded open set whose boundary is Ahlfors regular, then (5.4) holds for each p (1, ). Proof. Fix p (1, ). For f L p ( ), x, set (5.19) Ψf(x) = 1 ψ(x, y)f(y) dσ(y), V (x) where (5.20) ψ(x, y) = ( 1 x y ), ϕ(x) = dist(x, ), 2ϕ(x) +
19 19 and (5.21) V (x) = ψ(x, y) dσ(y). It is readily checked that (5.22) Ψ : C( ) C(), Ψf = f, and that (5.23) N Ψf CMf, f L 1 ( ), where Mf is the Hardy-Littlewood maximal function of f L 1 ( ), given that is Ahlfors regular. Hence (5.24) N Ψf L p ( ) C p f L p ( ), 1 < p <. We claim that for each p (1, ), (5.25) Ψ : L p ( ) L p, (Ψf) b = f σ-a.e. In light of (5.24), only the nontangential convergence of Ψf to f remains to be justified. This follows by taking f k C( ), f k f in L p -norm, using (5.22) on f k and (5.24) on f f k. Now, to establish (5.4), we argue as follows. Take v L p, and set w = Ψv b. By (5.25), w L p and w b = v b. Then take f k C( ) such that f k v b in L p ( )-norm, and set w k = Ψf k. By (5.22), each w k C(). By (5.24), (5.26) N ( w k w) L p ( ) C p f k v b L p ( ) 0, which is stronger than the L 1 estimate demanded in (5.4). As mentioned in the paragraph after (5.4), having such continuous functions suffices, since they are easily approximated by Lipschitz functions. Proposition 5.2 is proven. We now show that Ahlfors regularity implies (5.3). x set (5.27) Γ x = {y : x y 10 dist(y, )}. To be definite, for each Proposition 5.3. Let R m be a bounded open set with Ahlfors regular boundary. Then (5.3) holds. Proof. We first note that it suffices to prove that 1 (5.28) v dx C N v L1( ), v L δ Õ δ where Õδ = O δ \ O δ/2, since (5.3) then follows by applying (5.28) with δ replaced by 2 j δ and summing over j Z +. We now bring in the following lemma. 1,
20 20 Lemma 5.4. There exists K = K m with the following property. (0, (diam )/10], there exists a covering of by a collection For each δ (5.29) C = C 1 C K of balls of radius δ, centered in, such that for each k {1,..., K}, if B and B are distinct balls in C k, their centers are separated by a distance 10δ. We postpone the proof of Lemma 5.4, and show how it is used to prove Proposition 5.3. To begin, take a collection C = C 1 C K of balls of radius δ covering, with the properties stated above. Then C # = C # 1 C# K, consisting of balls concentric with those of C with radius 2δ, covers O δ. Furthermore, there exists A = A m (0, ) such that one can cover each ball B C # k by balls B 1,..., B A of radius δ/8, centered at points in B. Now form collections of balls C # kl, 1 k K, 1 l A, with each of the balls B 1,..., B A covering B C # k, described above, put into a different one of the collections C # kl. Throw away some balls from C# kl, thinning them out to a minimal collection (5.30) C = k K,l A covering Õδ. For each (k, l), any two distinct balls in C kl have centers separated by a distance 7δ. Each ball B C kl has radius δ/8 and each point p B has distance from lying between δ/4 and 5δ/4. For each such B, we will compare v dx with the integral of N v over a certain set Ã(B), which we proceed B to define. Given y, d(y) = dist(y, ), consider (5.31) A(y) = {x : y Γ x }. There exists p such that y p = d(y), and certainly p A(y). Also, if (5.27) holds, then (5.32) A(y) B 9d(y) (p). Now, for a ball B C kl, set (5.33) A(B) = {x : B Γ x }. If B is centered at y and p is closest to y, then (5.32) holds with d(y) 3δ/8. For each y B, d(y ) δ/4 and y p d(y) + δ/8. Now d(y) (9/8)δ, so y p (5/4)δ 5d(y ), and hence (5.34) A(B) B d(y) (p). C kl
21 21 We denote the right side of (5.34) by Ã(B). The use of Ã(B) in establishing (5.28) arises from the estimates (5.35) sup B v inf N v, and H m 1 (Ã(B)) C 1δ m 1, Ã(B) the latter estimate due to the hypothesis that is Ahlfors regular. Hence (5.36) 1 δ B v dx C m δ m 1 C Ã(B) inf N v Ã(B) N v dσ. Furthermore, the separation properties established for balls in each collection C kl yield (5.37) B B C kl = Ã(B) Ã(B ) =, so for each k K, l A, (5.38) 1 δ B C kl B v dx C N v dσ Ã(B),B C kl C N v L 1 ( ). Consequently, (5.39) 1 δ Õ δ v dx 1 δ A K l=1 k=1 B C kl B CAK N v L 1 ( ), v dx and Proposition 5.3 is established, modulo the proof of Lemma 5.4, to which we now turn. Proof of Lemma 5.4. Pick a maximal set {p k } of points that are separated by a distance δ. (Such clearly exists.) Set C = {B δ (p k )}. This collection covers, since if q were not covered one could add it to {p k }. On the other hand, the balls in B = {B δ/2 (p k )} are mutually disjoint balls in R m. Our task is to partition C as indicated in (5.29). Tile R m with cubes Q of edge 20 δ, edges parallel to the coordinate axes. Each such cube has a natural partition into 20 m little cubes (call them cells), of edge δ.
22 22 Label the cells in each Q from 1 to 20 m, in a fashion so that translating one such Q to the position of another preserves the labels of the cells. Note that for each cell C all the balls in B intersecting C lie in a cube of edge 2δ, so since they are mutually disjoint there can be at most [4 m /V m ] of them, where V m is the volume of the unit ball in R m. Now for each l {1,..., 20 m }, look at each cube Q and pick one ball from B (say B δ/2 (p k )) that intersects the cell labeled l in Q. (If no ball in B intersects the cell, pass it by.) Collect all these balls, as Q varies. Then take the corresponding balls B δ (p k ) in C. This forms a collection, which we will denote C l1, of balls from C, whose centers are separated by a distance 10δ. If C is not exhausted, do this again, to form C l2, for l {1,..., 20 m }, and repeat, forming C lk, k 3. All the balls in C will be exhausted by the time k = [4 m /V m ], so the lemma is established, with K m = 20 m [4 m /V m ]. Putting together Propositions , we have: Theorem 5.5. Assume is Ahlfors regular and satisfies (5.1). p > 1, (5.40) v L p, and div v L 1 (), If, for some then (5.41) div v dx = ν v b dσ.
23 23 6. Gauss-Green formulas on Riemannian manifolds Let R m carry a continuous metric tensor (g jk ), in addition to the Euclidean metric tensor (δ jk ). A vector field v = v j j has divergence div v D (R m ) given by (6.1) ϕ, div v = j ϕ, g 1/2 v j, where g = det(g jk ), and we use the summation convention. If div v is a locally integrable multiple of Lebesgue measure, or equivalently of dv = g 1/2 dx, we identify div v with div v dv. We denote by div 0 v this quantity associated to (δ jk ) rather than (g jk ), so div v = g 1/2 div 0 (g 1/2 v), in the locally integrable case. Let R m be an open set with locally finite perimeter. Assume v belongs to (6.2) D = {v C 0 0(R m, R m ) : div v L 1 (R m )} = {v C 0 0(R m, R m ) : div 0 (g 1/2 v) L 1 (R m )}. Then (6.3) div v dv = div 0 (g 1/2 v) dx. Hence Proposition 1.1 gives (6.4) div v dv = n v g 1/2 dσ, where n is the outward-pointing unit normal with respect to the metric (δ jk ) and σ is (m 1)-dimensional Hausdorff measure defined by (δ jk ). We claim that (6.5) n v g 1/2 dσ = ν, v g dσ g, where ν is the unit outward-pointing normal determined by (g jk ),, g is the inner product determined by (g jk ), and σ g is (m 1)-dimensional Hausdorff measure determined by (g jk ). The vectors ν = ν j j and n = n j j have asociated covectors (6.6) ν b = ν j dx j, n b = n j dx j, where (6.7) ν j = g jk ν k, n j = δ jk n k.
24 24 The covectors ν b and n b are parallel and both have unit length, with respect to their associated metric tensors, so (6.8) n j = aν j, a 2 = n b, n b g = g jk n j n k. Hence (6.9) n v g 1/2 = n j v j g 1/2 = g 1/2 n b, n b 1/2 g ν j v j = g 1/2 n b, n b 1/2 g ν, v g. Thus (6.5) is equivalent to the assertion that (6.10) σ g = g 1/2 n b, n b 1/2 g on measurable subsets of. The following result establishes (6.10). Proposition 6.1. Let S R m be a countably rectifiable (m 1)-dimensional set in R m, with measure-theoretic unit normal n determined by the Euclidean structure. If σ is (m 1)-dimensional Hausdorff measure determined by (δ jk ) and σ g is (m 1)- dimensional Hausdorff measure determined by (g jk ), then (6.10) holds on S. Proof. It is clear from the definitions that for each compact K R m there exists C K (1, ) such that (6.11) C 1 K σ(a) σ g(a) C K σ(a), A K. The hypothesis of countable rectifiability implies there is a disjoint union (6.12) S = k 1 M k N, σ where each M k is a Borel subset of some (m 1)-dimensional C 1 submanifold of R m, while σ(n) = 0, hence σ g (N) = 0. It is elementary that (6.10) holds on each C 1 submanifold of R m, of dimension m 1, so by (6.12) it holds on S. Since is countably rectifiable, we have the following variant of Proposition 1.1. Proposition 6.2. Let R m have a continuous metric tensor (g jk ). Let R m be an open set with locally finite perimeter. Then (6.13) div v dv = ν, v g dσ g, for all v D, defined by (6.2).
25 25 Similarly we can apply Proposition 1.2 to deduce that (6.14) div 0 (g 1/2 v) dx = n v g 1/2 dσ, whenever has a tame interior approximation and v belongs to (6.15) D = {v C 0 0(, R m ) : div 0 (g 1/2 v) L 1 ()} = {v C 0 0(, R m ) : div v L 1 ()}. We obtain Proposition 6.3. If R m has a continuous metric tensor (g jk ) and R m has a tame interior approximation, then (6.13) holds for all v D. Also the results of 5 together with Proposition 6.1 yield the following. Proposition 6.4. If R m has a continuous metric tensor (g jk ) and R m is a bounded domain with Ahlfors regular boundary, then (6.13) holds whenever (6.16) div v L 1 () and v L p, for some p > 1, where (6.17) L p = {v C() : N v L p ( ) and nontangential limit v b, σ-a.e.}. Remark. Using partitions of unity, we can extend the scope of these results to M, where M is a smooth manifold with a continuous metric tensor. We can apply Proposition 6.4 in the following setting. Let M be a compact manifold with a Riemannian metric whose components are continuous with a modulus of continuity ω satisfying (6.18) 1 0 ω(t) dt <. t Let V L (M) satisfy V 0 on M and V > 0 on a set of positive measure. Then let E(x, y) be the integral kernel of ( V ) 1 on L 2 (M). Let M be a connected NTA domain with Ahlfors regular boundary, or more generally a connected UR domain, as defined in [HMT]. For f L p ( ), set (6.19) Sf(x) = E(x, y)f(y) dσ g (y), x.
26 26 A fundamental result (established in [HMT]) is (6.20) N ( Sf) L p ( ) C p f L p ( ), 1 < p <, and that nontangential limits of Sf exist σ g -a.e. on. In particular, ( (6.21) lim ν(x) Sf(y) = 1 y x in Γ x 2 + K ) f(x), σ g -a.e., where K : L p ( ) L p ( ), for 1 < p <. Also, by Proposition 4.1, (6.22) Sf L p () + Sf L q () C f L 2 ( ) for some p, q > 2, and N Sf L r ( ) C f L 2 ( ) for some r > 2. Now if we take f L 2 ( ) and set (6.23) u = Sf, v = u u, we have (6.24) div v = u 2 + u u = u 2 + V u 2, on. Thus Proposition 6.4 applies to give (6.25) ( u 2 + V u 2 ) dv = ( u K ) f dσ g.
27 27 A. The Gauss-Green formula on smoothly bounded domains Our goal here is to give a simple, direct proof of the identity (A.1) div v dx = ν v dσ, for v C 0 (R m, R m ), when R m has the form (A.2) = {(x, x m ) R m : x m > A(x )}, with (A.3) A C 1 (R m ). In 2 it was shown how such an identity (actually for A C (R m )) plus an approximation argument yields such a result for domains of the form (A.2) with much rougher A, satisfying merely (2.2). We can deduce (A.1) from the following result. Proposition A.1. If R m satisfies (A.2) (A.3) and f C0 (R m ), and if e is an element of R m, then (A.4) e f(x) dx = (e ν)f dσ. In fact, taking {e j } to be the standard orthonormal basis of R m, replacing e by e j and f by v j (the jth component of v) in (A.4) and summing, we obtain (A.1). To begin the proof of (A.4), write (A.5) f x m dx = R m 1 ( x m >A(x ) = f(x, A(x )) dx R m 1 = (e m ν)f dσ. m f(x, x m ) dx m ) dx The first identity in (A.5) follows from Fubini s theorem, the second identity from the fundamental theorem of calculus, and the third identity from the formula (A.6) ν(x) = ( 1 + A(x ) 2) 1/2 ( A(x ), 1),
28 28 where x = (x, A(x )), which is the standard formula for the downward pointing normal to the C 1 surface x m = A(x ), and from the formula (A.7) dσ(x) = 1 + A(x ) 2 dx, for the surface area of the graph of a C 1 function. This establishes (A.4) for e = e m. Here are some useful comments regarding the surface area formula (A.7). The (m 1)-dimensional surface R m inherits a Riemannian metric tensor. A coordinate system on a Riemannian manifold M such as produces components (g jk ) for the metric tensor, and the area element is given by (A.8) dσ(y) = g(y) dy, in a coordinate system y = (y 1,..., y m 1 ), where g(y) = det(g jk (y)). In case the coordinate system on = {x m = A(x )} is x (x, A(x )), the formula (A.8) specializes to (A.7). It is also useful to note that the transformation properties for metric tensors and for integrals under C 1 diffeomorphisms imply that (A.8) defines area on a Riemannian manifold in a coordinate invariant fashion. To continue the proof of (A.4), since v has compact support, we can assume A(x ) L for all x, for some L <. Then has a representation of the form (A.2) (A.3) in new coordinates obtained by any rotation sufficiently close to the identity. Hence the identity (A.4) holds when e = e m is replaced by any sufficiently close element of R m. In particular, it works for e = e m + ae j, for 1 j m 1 and for a sufficiently small. Thus we have (A.9) (e m + ae j ) f(x) dx = (e m + ae j ) ν f dσ. If we subtract (A.5) from this and divide the result by a, we obtain (A.4) for e = e j, for all j, and hence (A.4) holds in general. This completes the proof. Remark 1. Using again (A.6) (A.7), we can rewrite (A.1) as (A.10) div v dx = which is the form used in (2.11). R m 1 ( A(x ), 1) v(x, A(x )) dx, Remark 2. For a C 1 graph, or more generally for a C 1 manifold with a continuous metric tensor, one can show that (A.7) and (A.8) coincide with Hausdorff measure, directly from the definitions and the fact that H m 1 coincides with Lebesgue measure on R m 1. Details are given in [T], Propositions 12.6 and 12.7.
29 29 B. Variant of a result of Chen and Torres Let O R m be open, and set (B.1) D = {v L (O, R m ) : div v L 1 (O)}. This is a Banach space with norm (B.2) v D = v L (O) + div v L 1 (O). Let O be an open set with finite perimeter, and take σ and ν as in (1.4) (1.5). The following result was established in [CT]. Proposition B.1. There is a continuous linear map (called the normal trace) (B.3) τ ν : D L (, σ) such that, for each ϕ Lip(O), (B.4) ϕ div v dx + ϕ v dx = τ ν (v)ϕ dσ. Furthermore, one has (B.5) τ ν v L (,σ) lim δ 0 v L (U δ ), where U δ = {x : dist(x, ) δ}, and (B.6) v D C(O) = τ ν v = ν v. In fact, [CT] has a stronger result, allowing div v to be a finite measure on O. Note that Proposition B.1 extends Proposition 1.1, but it does not imply Proposition 1.2 or the results of 5. Proof of Proposition B.1: Pick ψ C0 (R m ) such that ψ dx = 1, ψ 0, and ψ(x) = 0 for x 1, and set ψ k (x) = k m ψ(kx). Given v D, set v k = ψ k v, which is well defined and smooth on a neighborhood of for large k. For each ϕ Lip(O), ϕv k is Lipschitz on a neighborhood of and div(ϕv k ) = ϕ div v k + ϕ v k, so (1.4) gives (B.7) ϕ div v k dx + ϕ v k dx = (ν v k )ϕ dσ.
30 30 Now the left side of (B.7) converges to (B.8) ϕ div v dx + ϕ v dx = L v ϕ, as k, so the right side also converges: (B.9) (ν v k )ϕ dσ L v ϕ. We have L v : Lip(O) C for each v D, and (B.9) gives the estimate (B.10) L v ϕ lim sup k sup v k (x) ϕ L 1 (,σ) lim δ 0 v L (U δ ) ϕ L 1 (,σ). Since ϕ ϕ maps Lip(O) onto a dense linear subspace of L 1 (, σ), we see that L v has a unique extension to a continuous linear functional on L 1 (, σ), with norm lim δ 0 v L (U δ ). Since L 1 (, σ) = L (, σ), we have τ ν (v) L (, σ) such that (B.5) holds and (B.11) L v ϕ = τ ν (v)ϕ dσ. Comparison with (B.8) gives (B.4). Finally, under the hypothesis in (B.6) we clearly have v k v uniformly on, so (B.6) follows from (B.9). Remark. Note that D in (B.1) is a module over C 0 (O). Hence there is no loss of generality in replacing O by R m in (B.1).
31 31 References [CT] G.-Q. Chen and M. Torres, Divergence-measure fields, sets of finite perimeter, and conservation laws, Arch. Rational Mech. Anal. 175 (2005), [CTZ] G.-Q. Chen, M. Torres, and W. Ziemer, Measure-theoretical analysis and nonlinear conservation laws, Preprint, May [EG] L. Evans and R. Gariepy, Measure Theory and Fine Properties of Functions, CRC Press, [Fed1] H. Federer, Measure and area, Bull. AMS 58 (1952), [Fed2] H. Federer, The area of a nonparametric surface, Proc. AMS 11 (1960), [Fed3] H. Federer, Geometric Measure Theory, Springer-Verlag, New York, [HMT] S. Hofmann, M. Mitrea, and M. Taylor, Singular integrals and elliptic boundary problems on regular Semmes-Kenig-Toro domains, IMRN 2010 (2010), [T] M. Taylor, Measure Theory and Integration, Graduate Studies in Math., American Mathematical Society, Providence, RI, [Tom] R. Tompson, Area of k-dimensional nonparametric surfaces in k + 1 space, Trans. AMS 77 (1954), [Zie] W. Ziemer, Weakly Differentiable Functions, Springer-Verlag, New York, 1989.
The Gauss-Green Formula (And Elliptic Boundary Problems On Rough Domains) Joint Work with Steve Hofmann and Marius Mitrea
The Gauss-Green Formula And Elliptic Boundary Problems On Rough Domains) Joint Work with Steve Hofmann and Marius Mitrea Dirichlet Problem on open in a compact Riemannian manifold M), dimension n: ) Lu
More informationSingular Integrals and Elliptic Boundary Problems on Regular Semmes Kenig Toro Domains
S. Hofmann et al. (2010 Singular Integrals and Elliptic Boundary Problems on Regular Semmes Kenig Toro Domains, International Mathematics Research Notices, Vol. 2010, No. 14, pp. 2567 2865 Advance Access
More informationSingular integrals and elliptic boundary problems on regular Semmes-Kenig-Toro domains
Singular integrals and elliptic boundary problems on regular Semmes-Kenig-Toro domains Steve Hofmann Mathematics Department, University of Missouri Marius Mitrea Mathematics Department, University of Missouri,
More informationRecall that if X is a compact metric space, C(X), the space of continuous (real-valued) functions on X, is a Banach space with the norm
Chapter 13 Radon Measures Recall that if X is a compact metric space, C(X), the space of continuous (real-valued) functions on X, is a Banach space with the norm (13.1) f = sup x X f(x). We want to identify
More informationThe oblique derivative problem for general elliptic systems in Lipschitz domains
M. MITREA The oblique derivative problem for general elliptic systems in Lipschitz domains Let M be a smooth, oriented, connected, compact, boundaryless manifold of real dimension m, and let T M and T
More informationCHAPTER 6. Differentiation
CHPTER 6 Differentiation The generalization from elementary calculus of differentiation in measure theory is less obvious than that of integration, and the methods of treating it are somewhat involved.
More informationPotential Analysis meets Geometric Measure Theory
Potential Analysis meets Geometric Measure Theory T. Toro Abstract A central question in Potential Theory is the extend to which the geometry of a domain influences the boundary regularity of the solution
More informationHerz (cf. [H], and also [BS]) proved that the reverse inequality is also true, that is,
REARRANGEMENT OF HARDY-LITTLEWOOD MAXIMAL FUNCTIONS IN LORENTZ SPACES. Jesús Bastero*, Mario Milman and Francisco J. Ruiz** Abstract. For the classical Hardy-Littlewood maximal function M f, a well known
More informationGeometric intuition: from Hölder spaces to the Calderón-Zygmund estimate
Geometric intuition: from Hölder spaces to the Calderón-Zygmund estimate A survey of Lihe Wang s paper Michael Snarski December 5, 22 Contents Hölder spaces. Control on functions......................................2
More informationNECESSARY CONDITIONS FOR WEIGHTED POINTWISE HARDY INEQUALITIES
NECESSARY CONDITIONS FOR WEIGHTED POINTWISE HARDY INEQUALITIES JUHA LEHRBÄCK Abstract. We establish necessary conditions for domains Ω R n which admit the pointwise (p, β)-hardy inequality u(x) Cd Ω(x)
More informationGIOVANNI COMI AND MONICA TORRES
ONE-SIDED APPROXIMATION OF SETS OF FINITE PERIMETER GIOVANNI COMI AND MONICA TORRES Abstract. In this note we present a new proof of a one-sided approximation of sets of finite perimeter introduced in
More informationand BV loc R N ; R d)
Necessary and sufficient conditions for the chain rule in W 1,1 loc R N ; R d) and BV loc R N ; R d) Giovanni Leoni Massimiliano Morini July 25, 2005 Abstract In this paper we prove necessary and sufficient
More information2. Function spaces and approximation
2.1 2. Function spaces and approximation 2.1. The space of test functions. Notation and prerequisites are collected in Appendix A. Let Ω be an open subset of R n. The space C0 (Ω), consisting of the C
More informationSobolev Spaces. Chapter 10
Chapter 1 Sobolev Spaces We now define spaces H 1,p (R n ), known as Sobolev spaces. For u to belong to H 1,p (R n ), we require that u L p (R n ) and that u have weak derivatives of first order in L p
More informationSOLUTION OF THE DIRICHLET PROBLEM WITH L p BOUNDARY CONDITION. Dagmar Medková
29 Kragujevac J. Math. 31 (2008) 29 42. SOLUTION OF THE DIRICHLET PROBLEM WITH L p BOUNDARY CONDITION Dagmar Medková Czech Technical University, Faculty of Mechanical Engineering, Department of Technical
More informationSHARP BOUNDARY TRACE INEQUALITIES. 1. Introduction
SHARP BOUNDARY TRACE INEQUALITIES GILES AUCHMUTY Abstract. This paper describes sharp inequalities for the trace of Sobolev functions on the boundary of a bounded region R N. The inequalities bound (semi-)norms
More informationON BOUNDEDNESS OF MAXIMAL FUNCTIONS IN SOBOLEV SPACES
Annales Academiæ Scientiarum Fennicæ Mathematica Volumen 29, 2004, 167 176 ON BOUNDEDNESS OF MAXIMAL FUNCTIONS IN SOBOLEV SPACES Piotr Haj lasz and Jani Onninen Warsaw University, Institute of Mathematics
More informationLaplace s Equation. Chapter Mean Value Formulas
Chapter 1 Laplace s Equation Let be an open set in R n. A function u C 2 () is called harmonic in if it satisfies Laplace s equation n (1.1) u := D ii u = 0 in. i=1 A function u C 2 () is called subharmonic
More information3 (Due ). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure?
MA 645-4A (Real Analysis), Dr. Chernov Homework assignment 1 (Due ). Show that the open disk x 2 + y 2 < 1 is a countable union of planar elementary sets. Show that the closed disk x 2 + y 2 1 is a countable
More informationThe optimal partial transport problem
The optimal partial transport problem Alessio Figalli Abstract Given two densities f and g, we consider the problem of transporting a fraction m [0, min{ f L 1, g L 1}] of the mass of f onto g minimizing
More information+ 2x sin x. f(b i ) f(a i ) < ɛ. i=1. i=1
Appendix To understand weak derivatives and distributional derivatives in the simplest context of functions of a single variable, we describe without proof some results from real analysis (see [7] and
More information2 (Bonus). Let A X consist of points (x, y) such that either x or y is a rational number. Is A measurable? What is its Lebesgue measure?
MA 645-4A (Real Analysis), Dr. Chernov Homework assignment 1 (Due 9/5). Prove that every countable set A is measurable and µ(a) = 0. 2 (Bonus). Let A consist of points (x, y) such that either x or y is
More informationFUNCTIONS OF BOUNDED VARIATION AND SETS OF FINITE PERIMETER
FUNCTIONS OF BOUNDED VARIATION AND SETS OF FINITE PERIMETER MONICA TORRES Abstract. We present in these notes some fine properties of functions of bounded variation and sets of finite perimeter, which
More informationBehaviour of Lipschitz functions on negligible sets. Non-differentiability in R. Outline
Behaviour of Lipschitz functions on negligible sets G. Alberti 1 M. Csörnyei 2 D. Preiss 3 1 Università di Pisa 2 University College London 3 University of Warwick Lars Ahlfors Centennial Celebration Helsinki,
More informationCoupled second order singular perturbations for phase transitions
Coupled second order singular perturbations for phase transitions CMU 06/09/11 Ana Cristina Barroso, Margarida Baía, Milena Chermisi, JM Introduction Let Ω R d with Lipschitz boundary ( container ) and
More informationThe Caratheodory Construction of Measures
Chapter 5 The Caratheodory Construction of Measures Recall how our construction of Lebesgue measure in Chapter 2 proceeded from an initial notion of the size of a very restricted class of subsets of R,
More informationBIHARMONIC WAVE MAPS INTO SPHERES
BIHARMONIC WAVE MAPS INTO SPHERES SEBASTIAN HERR, TOBIAS LAMM, AND ROLAND SCHNAUBELT Abstract. A global weak solution of the biharmonic wave map equation in the energy space for spherical targets is constructed.
More informationTHEOREMS, ETC., FOR MATH 516
THEOREMS, ETC., FOR MATH 516 Results labeled Theorem Ea.b.c (or Proposition Ea.b.c, etc.) refer to Theorem c from section a.b of Evans book (Partial Differential Equations). Proposition 1 (=Proposition
More informationLebesgue Measure on R n
CHAPTER 2 Lebesgue Measure on R n Our goal is to construct a notion of the volume, or Lebesgue measure, of rather general subsets of R n that reduces to the usual volume of elementary geometrical sets
More informationExtension and Representation of Divergence-free Vector Fields on Bounded Domains. Tosio Kato, Marius Mitrea, Gustavo Ponce, and Michael Taylor
Extension and Representation of Divergence-free Vector Fields on Bounded Domains Tosio Kato, Marius Mitrea, Gustavo Ponce, and Michael Taylor 1. Introduction Let Ω R n be a bounded, connected domain, with
More informationReal Analysis Problems
Real Analysis Problems Cristian E. Gutiérrez September 14, 29 1 1 CONTINUITY 1 Continuity Problem 1.1 Let r n be the sequence of rational numbers and Prove that f(x) = 1. f is continuous on the irrationals.
More informationA RELATIONSHIP BETWEEN THE DIRICHLET AND REGULARITY PROBLEMS FOR ELLIPTIC EQUATIONS. Zhongwei Shen
A RELATIONSHIP BETWEEN THE DIRICHLET AND REGULARITY PROBLEMS FOR ELLIPTIC EQUATIONS Zhongwei Shen Abstract. Let L = diva be a real, symmetric second order elliptic operator with bounded measurable coefficients.
More informationTHE DIRICHLET PROBLEM WITH BM O BOUNDARY DATA AND ALMOST-REAL COEFFICIENTS
THE DIRICHLET PROBLEM WITH BM O BOUNDARY DATA AND ALMOST-REAL COEFFICIENTS ARIEL BARTON Abstract. It is known that a function, harmonic in a Lipschitz domain, is the Poisson extension of a BMO function
More informationREAL ANALYSIS I HOMEWORK 4
REAL ANALYSIS I HOMEWORK 4 CİHAN BAHRAN The questions are from Stein and Shakarchi s text, Chapter 2.. Given a collection of sets E, E 2,..., E n, construct another collection E, E 2,..., E N, with N =
More informationEXPOSITORY NOTES ON DISTRIBUTION THEORY, FALL 2018
EXPOSITORY NOTES ON DISTRIBUTION THEORY, FALL 2018 While these notes are under construction, I expect there will be many typos. The main reference for this is volume 1 of Hörmander, The analysis of liner
More informationHARMONIC ANALYSIS. Date:
HARMONIC ANALYSIS Contents. Introduction 2. Hardy-Littlewood maximal function 3. Approximation by convolution 4. Muckenhaupt weights 4.. Calderón-Zygmund decomposition 5. Fourier transform 6. BMO (bounded
More informationChapter One. The Calderón-Zygmund Theory I: Ellipticity
Chapter One The Calderón-Zygmund Theory I: Ellipticity Our story begins with a classical situation: convolution with homogeneous, Calderón- Zygmund ( kernels on R n. Let S n 1 R n denote the unit sphere
More informationarxiv: v1 [math.ca] 7 Aug 2015
THE WHITNEY EXTENSION THEOREM IN HIGH DIMENSIONS ALAN CHANG arxiv:1508.01779v1 [math.ca] 7 Aug 2015 Abstract. We prove a variant of the standard Whitney extension theorem for C m (R n ), in which the norm
More informationBoot camp - Problem set
Boot camp - Problem set Luis Silvestre September 29, 2017 In the summer of 2017, I led an intensive study group with four undergraduate students at the University of Chicago (Matthew Correia, David Lind,
More informationIntegration on Measure Spaces
Chapter 3 Integration on Measure Spaces In this chapter we introduce the general notion of a measure on a space X, define the class of measurable functions, and define the integral, first on a class of
More informationJUHA KINNUNEN. Harmonic Analysis
JUHA KINNUNEN Harmonic Analysis Department of Mathematics and Systems Analysis, Aalto University 27 Contents Calderón-Zygmund decomposition. Dyadic subcubes of a cube.........................2 Dyadic cubes
More informationLocal semiconvexity of Kantorovich potentials on non-compact manifolds
Local semiconvexity of Kantorovich potentials on non-compact manifolds Alessio Figalli, Nicola Gigli Abstract We prove that any Kantorovich potential for the cost function c = d / on a Riemannian manifold
More informationESTIMATES FOR THE MONGE-AMPERE EQUATION
GLOBAL W 2,p ESTIMATES FOR THE MONGE-AMPERE EQUATION O. SAVIN Abstract. We use a localization property of boundary sections for solutions to the Monge-Ampere equation obtain global W 2,p estimates under
More informationON THE REGULARITY OF SAMPLE PATHS OF SUB-ELLIPTIC DIFFUSIONS ON MANIFOLDS
Bendikov, A. and Saloff-Coste, L. Osaka J. Math. 4 (5), 677 7 ON THE REGULARITY OF SAMPLE PATHS OF SUB-ELLIPTIC DIFFUSIONS ON MANIFOLDS ALEXANDER BENDIKOV and LAURENT SALOFF-COSTE (Received March 4, 4)
More informationMeasurable Choice Functions
(January 19, 2013) Measurable Choice Functions Paul Garrett garrett@math.umn.edu http://www.math.umn.edu/ garrett/ [This document is http://www.math.umn.edu/ garrett/m/fun/choice functions.pdf] This note
More informationHOMEOMORPHISMS OF BOUNDED VARIATION
HOMEOMORPHISMS OF BOUNDED VARIATION STANISLAV HENCL, PEKKA KOSKELA AND JANI ONNINEN Abstract. We show that the inverse of a planar homeomorphism of bounded variation is also of bounded variation. In higher
More informationTHE HARDY LITTLEWOOD MAXIMAL FUNCTION OF A SOBOLEV FUNCTION. Juha Kinnunen. 1 f(y) dy, B(x, r) B(x,r)
Appeared in Israel J. Math. 00 (997), 7 24 THE HARDY LITTLEWOOD MAXIMAL FUNCTION OF A SOBOLEV FUNCTION Juha Kinnunen Abstract. We prove that the Hardy Littlewood maximal operator is bounded in the Sobolev
More informationDifferentiation of Measures and Functions
Chapter 6 Differentiation of Measures and Functions This chapter is concerned with the differentiation theory of Radon measures. In the first two sections we introduce the Radon measures and discuss two
More information3. (a) What is a simple function? What is an integrable function? How is f dµ defined? Define it first
Math 632/6321: Theory of Functions of a Real Variable Sample Preinary Exam Questions 1. Let (, M, µ) be a measure space. (a) Prove that if µ() < and if 1 p < q
More informationExistence of minimizers for the pure displacement problem in nonlinear elasticity
Existence of minimizers for the pure displacement problem in nonlinear elasticity Cristinel Mardare Université Pierre et Marie Curie - Paris 6, Laboratoire Jacques-Louis Lions, Paris, F-75005 France Abstract
More informationABSOLUTE CONTINUITY OF FOLIATIONS
ABSOLUTE CONTINUITY OF FOLIATIONS C. PUGH, M. VIANA, A. WILKINSON 1. Introduction In what follows, U is an open neighborhood in a compact Riemannian manifold M, and F is a local foliation of U. By this
More informationTHE L 2 -HODGE THEORY AND REPRESENTATION ON R n
THE L 2 -HODGE THEORY AND REPRESENTATION ON R n BAISHENG YAN Abstract. We present an elementary L 2 -Hodge theory on whole R n based on the minimization principle of the calculus of variations and some
More informationTHE RADON NIKODYM PROPERTY AND LIPSCHITZ EMBEDDINGS
THE RADON NIKODYM PROPERTY AND LIPSCHITZ EMBEDDINGS Motivation The idea here is simple. Suppose we have a Lipschitz homeomorphism f : X Y where X and Y are Banach spaces, namely c 1 x y f (x) f (y) c 2
More informationNOTES ON THE REGULARITY OF QUASICONFORMAL HOMEOMORPHISMS
NOTES ON THE REGULARITY OF QUASICONFORMAL HOMEOMORPHISMS CLARK BUTLER. Introduction The purpose of these notes is to give a self-contained proof of the following theorem, Theorem.. Let f : S n S n be a
More informationFunctional Analysis HW #3
Functional Analysis HW #3 Sangchul Lee October 26, 2015 1 Solutions Exercise 2.1. Let D = { f C([0, 1]) : f C([0, 1])} and define f d = f + f. Show that D is a Banach algebra and that the Gelfand transform
More informationSobolev spaces. May 18
Sobolev spaces May 18 2015 1 Weak derivatives The purpose of these notes is to give a very basic introduction to Sobolev spaces. More extensive treatments can e.g. be found in the classical references
More informationVector Integration and Disintegration of Measures. Michael Taylor
Vector Integration and Disintegration of Measures Michael Taylor 1. Introduction Let and be compact metric spaces, and π : a continuous map of onto. Let µ be a positive, finite Borel measure on. The following
More informationFriedrich symmetric systems
viii CHAPTER 8 Friedrich symmetric systems In this chapter, we describe a theory due to Friedrich [13] for positive symmetric systems, which gives the existence and uniqueness of weak solutions of boundary
More informationSolutions to Problem Set 5 for , Fall 2007
Solutions to Problem Set 5 for 18.101, Fall 2007 1 Exercise 1 Solution For the counterexample, let us consider M = (0, + ) and let us take V = on M. x Let W be the vector field on M that is identically
More informationPeak Point Theorems for Uniform Algebras on Smooth Manifolds
Peak Point Theorems for Uniform Algebras on Smooth Manifolds John T. Anderson and Alexander J. Izzo Abstract: It was once conjectured that if A is a uniform algebra on its maximal ideal space X, and if
More informationChapter 2 Metric Spaces
Chapter 2 Metric Spaces The purpose of this chapter is to present a summary of some basic properties of metric and topological spaces that play an important role in the main body of the book. 2.1 Metrics
More informationWiener Measure and Brownian Motion
Chapter 16 Wiener Measure and Brownian Motion Diffusion of particles is a product of their apparently random motion. The density u(t, x) of diffusing particles satisfies the diffusion equation (16.1) u
More informationKUIPER S THEOREM ON CONFORMALLY FLAT MANIFOLDS
KUIPER S THEOREM ON CONFORMALLY FLAT MANIFOLDS RALPH HOWARD DEPARTMENT OF MATHEMATICS UNIVERSITY OF SOUTH CAROLINA COLUMBIA, S.C. 29208, USA HOWARD@MATH.SC.EDU 1. Introduction These are notes to that show
More informationReal Analysis Notes. Thomas Goller
Real Analysis Notes Thomas Goller September 4, 2011 Contents 1 Abstract Measure Spaces 2 1.1 Basic Definitions........................... 2 1.2 Measurable Functions........................ 2 1.3 Integration..............................
More informationWavelets and modular inequalities in variable L p spaces
Wavelets and modular inequalities in variable L p spaces Mitsuo Izuki July 14, 2007 Abstract The aim of this paper is to characterize variable L p spaces L p( ) (R n ) using wavelets with proper smoothness
More informationMaths 212: Homework Solutions
Maths 212: Homework Solutions 1. The definition of A ensures that x π for all x A, so π is an upper bound of A. To show it is the least upper bound, suppose x < π and consider two cases. If x < 1, then
More informationCones of measures. Tatiana Toro. University of Washington. Quantitative and Computational Aspects of Metric Geometry
Cones of measures Tatiana Toro University of Washington Quantitative and Computational Aspects of Metric Geometry Based on joint work with C. Kenig and D. Preiss Tatiana Toro (University of Washington)
More informationfor all x,y [a,b]. The Lipschitz constant of f is the infimum of constants C with this property.
viii 3.A. FUNCTIONS 77 Appendix In this appendix, we describe without proof some results from real analysis which help to understand weak and distributional derivatives in the simplest context of functions
More informationApplications of the periodic unfolding method to multi-scale problems
Applications of the periodic unfolding method to multi-scale problems Doina Cioranescu Université Paris VI Santiago de Compostela December 14, 2009 Periodic unfolding method and multi-scale problems 1/56
More informationPseudo-Poincaré Inequalities and Applications to Sobolev Inequalities
Pseudo-Poincaré Inequalities and Applications to Sobolev Inequalities Laurent Saloff-Coste Abstract Most smoothing procedures are via averaging. Pseudo-Poincaré inequalities give a basic L p -norm control
More informationMINIMAL GRAPHS PART I: EXISTENCE OF LIPSCHITZ WEAK SOLUTIONS TO THE DIRICHLET PROBLEM WITH C 2 BOUNDARY DATA
MINIMAL GRAPHS PART I: EXISTENCE OF LIPSCHITZ WEAK SOLUTIONS TO THE DIRICHLET PROBLEM WITH C 2 BOUNDARY DATA SPENCER HUGHES In these notes we prove that for any given smooth function on the boundary of
More informationMATH 411 NOTES (UNDER CONSTRUCTION)
MATH 411 NOTES (NDE CONSTCTION 1. Notes on compact sets. This is similar to ideas you learned in Math 410, except open sets had not yet been defined. Definition 1.1. K n is compact if for every covering
More information1 Cheeger differentiation
1 Cheeger differentiation after J. Cheeger [1] and S. Keith [3] A summary written by John Mackay Abstract We construct a measurable differentiable structure on any metric measure space that is doubling
More informationMAT 257, Handout 13: December 5-7, 2011.
MAT 257, Handout 13: December 5-7, 2011. The Change of Variables Theorem. In these notes, I try to make more explicit some parts of Spivak s proof of the Change of Variable Theorem, and to supply most
More informationON THE INTERPLAY BETWEEN SEVERAL COMPLEX VARIABLES, GEOMETRIC MEASURE THEORY, AND HARMONIC ANALYSIS
Dedicated with gratitude and respect to Professor Cabiria Andreian Cazacu ON THE INTERPLAY BETWEEN SEVERAL COMPLEX VARIABLES, GEOMETRIC MEASURE THEORY, AND HARMONIC ANALYSIS IRINA MITREA, MARIUS MITREA
More informationP(E t, Ω)dt, (2) 4t has an advantage with respect. to the compactly supported mollifiers, i.e., the function W (t)f satisfies a semigroup law:
Introduction Functions of bounded variation, usually denoted by BV, have had and have an important role in several problems of calculus of variations. The main features that make BV functions suitable
More informationPARTIAL REGULARITY OF BRENIER SOLUTIONS OF THE MONGE-AMPÈRE EQUATION
PARTIAL REGULARITY OF BRENIER SOLUTIONS OF THE MONGE-AMPÈRE EQUATION ALESSIO FIGALLI AND YOUNG-HEON KIM Abstract. Given Ω, Λ R n two bounded open sets, and f and g two probability densities concentrated
More information8 Singular Integral Operators and L p -Regularity Theory
8 Singular Integral Operators and L p -Regularity Theory 8. Motivation See hand-written notes! 8.2 Mikhlin Multiplier Theorem Recall that the Fourier transformation F and the inverse Fourier transformation
More informationFragmentability and σ-fragmentability
F U N D A M E N T A MATHEMATICAE 143 (1993) Fragmentability and σ-fragmentability by J. E. J a y n e (London), I. N a m i o k a (Seattle) and C. A. R o g e r s (London) Abstract. Recent work has studied
More informationA NOTE ON WEAK CONVERGENCE OF SINGULAR INTEGRALS IN METRIC SPACES
A NOTE ON WEAK CONVERGENCE OF SINGULAR INTEGRALS IN METRIC SPACES VASILIS CHOUSIONIS AND MARIUSZ URBAŃSKI Abstract. We prove that in any metric space (X, d) the singular integral operators Tµ,ε(f)(x) k
More informationON MAPS WHOSE DISTRIBUTIONAL JACOBIAN IS A MEASURE
[version: December 15, 2007] Real Analysis Exchange Summer Symposium 2007, pp. 153-162 Giovanni Alberti, Dipartimento di Matematica, Università di Pisa, largo Pontecorvo 5, 56127 Pisa, Italy. Email address:
More informationMATHS 730 FC Lecture Notes March 5, Introduction
1 INTRODUCTION MATHS 730 FC Lecture Notes March 5, 2014 1 Introduction Definition. If A, B are sets and there exists a bijection A B, they have the same cardinality, which we write as A, #A. If there exists
More informationLebesgue Measure on R n
8 CHAPTER 2 Lebesgue Measure on R n Our goal is to construct a notion of the volume, or Lebesgue measure, of rather general subsets of R n that reduces to the usual volume of elementary geometrical sets
More informationNotes on Distributions
Notes on Distributions Functional Analysis 1 Locally Convex Spaces Definition 1. A vector space (over R or C) is said to be a topological vector space (TVS) if it is a Hausdorff topological space and the
More informationLecture 11 Hyperbolicity.
Lecture 11 Hyperbolicity. 1 C 0 linearization near a hyperbolic point 2 invariant manifolds Hyperbolic linear maps. Let E be a Banach space. A linear map A : E E is called hyperbolic if we can find closed
More informationDEGREE AND SOBOLEV SPACES. Haïm Brezis Yanyan Li Petru Mironescu Louis Nirenberg. Introduction
Topological Methods in Nonlinear Analysis Journal of the Juliusz Schauder Center Volume 13, 1999, 181 190 DEGREE AND SOBOLEV SPACES Haïm Brezis Yanyan Li Petru Mironescu Louis Nirenberg Dedicated to Jürgen
More informationQuantitative Homogenization of Elliptic Operators with Periodic Coefficients
Quantitative Homogenization of Elliptic Operators with Periodic Coefficients Zhongwei Shen Abstract. These lecture notes introduce the quantitative homogenization theory for elliptic partial differential
More informationIntroduction to Geometric Measure Theory
Introduction to Geometric easure Theory Leon Simon 1 Leon Simon 2014 1 The research described here was partially supported by NSF grants DS-9504456 & DS 9207704 at Stanford University Contents 1 Preliminary
More informationSolutions: Problem Set 4 Math 201B, Winter 2007
Solutions: Problem Set 4 Math 2B, Winter 27 Problem. (a Define f : by { x /2 if < x
More information08a. Operators on Hilbert spaces. 1. Boundedness, continuity, operator norms
(February 24, 2017) 08a. Operators on Hilbert spaces Paul Garrett garrett@math.umn.edu http://www.math.umn.edu/ garrett/ [This document is http://www.math.umn.edu/ garrett/m/real/notes 2016-17/08a-ops
More informationESTIMATES FOR ELLIPTIC HOMOGENIZATION PROBLEMS IN NONSMOOTH DOMAINS. Zhongwei Shen
W,p ESTIMATES FOR ELLIPTIC HOMOGENIZATION PROBLEMS IN NONSMOOTH DOMAINS Zhongwei Shen Abstract. Let L = div`a` x, > be a family of second order elliptic operators with real, symmetric coefficients on a
More informationThe Dirichlet boundary problems for second order parabolic operators satisfying a Carleson condition
The Dirichlet boundary problems for second order parabolic operators satisfying a Carleson condition Sukjung Hwang CMAC, Yonsei University Collaboration with M. Dindos and M. Mitrea The 1st Meeting of
More informationMid Term-1 : Practice problems
Mid Term-1 : Practice problems These problems are meant only to provide practice; they do not necessarily reflect the difficulty level of the problems in the exam. The actual exam problems are likely to
More informationMath The Laplacian. 1 Green s Identities, Fundamental Solution
Math. 209 The Laplacian Green s Identities, Fundamental Solution Let be a bounded open set in R n, n 2, with smooth boundary. The fact that the boundary is smooth means that at each point x the external
More informationContinuous Functions on Metric Spaces
Continuous Functions on Metric Spaces Math 201A, Fall 2016 1 Continuous functions Definition 1. Let (X, d X ) and (Y, d Y ) be metric spaces. A function f : X Y is continuous at a X if for every ɛ > 0
More informationSobolev Mappings between Manifolds and Metric Spaces
Sobolev Mappings between Manifolds and Metric Spaces Piotr Haj lasz Abstract In connection with the theory of p-harmonic mappings, Eells and Lemaire raised a question about density of smooth mappings in
More informationn [ F (b j ) F (a j ) ], n j=1(a j, b j ] E (4.1)
1.4. CONSTRUCTION OF LEBESGUE-STIELTJES MEASURES In this section we shall put to use the Carathéodory-Hahn theory, in order to construct measures with certain desirable properties first on the real line
More informationWave equation on manifolds and finite speed of propagation
Wave equation on manifolds and finite speed of propagation Ethan Y. Jaffe Let M be a Riemannian manifold (without boundary), and let be the (negative of) the Laplace-Beltrami operator. In this note, we
More informationTHE INVERSE FUNCTION THEOREM FOR LIPSCHITZ MAPS
THE INVERSE FUNCTION THEOREM FOR LIPSCHITZ MAPS RALPH HOWARD DEPARTMENT OF MATHEMATICS UNIVERSITY OF SOUTH CAROLINA COLUMBIA, S.C. 29208, USA HOWARD@MATH.SC.EDU Abstract. This is an edited version of a
More informationSobolev Spaces. Chapter Hölder spaces
Chapter 2 Sobolev Spaces Sobolev spaces turn out often to be the proper setting in which to apply ideas of functional analysis to get information concerning partial differential equations. Here, we collect
More information