KATO S INEQUALITY UP TO THE BOUNDARY. Contents 1. Introduction 1 2. Properties of functions in X 4 3. Proof of Theorem
|
|
- Corey Chase
- 7 years ago
- Views:
Transcription
1 KATO S INEQUALITY UP TO THE BOUNDARY HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) Abstract. We show that if u is a finite measure in then, under suitable assumptions on u near, u + is also a finite measure in. We also study properties of the normal derivatives + and on. Contents 1. Introduction 1 2. Properties of functions in X 4 3. Proof of Theorem Properties of 8 5. Proof of Theorem Kato s inequality up to the boundary Computing + for W 2,1 -functions 16 Appendix A. The measure u + need not be finite 18 Appendix B. Approximation by smooth functions in 19 Appendix C. Proof of Lemma References Introduction Let R N be a smooth bounded domain. Given u L 1 () with u L 1 (), Kato s inequality (see [9]; see also [4]) asserts that (1.1) u + χ u in D (). In particular, (1.1) implies that u + is a locally finite measure in. Our goal in this paper is to address the question whether u + is a finite measure up to the boundary of, i.e., whether u + <. In general, the answer is negative: one can even construct harmonic functions u C() H 1 () such that u + is not a finite measure in ; see Proposition A.1 below. With further assumptions on u (for instance if u W 2,1 () or if u vanishes on the boundary) we will see that the answer is positive. Date: November 26,
2 2 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) The following class of functions will play a central role. We say that u X if u W 1,1 () and if there exists a constant C > 0 such that (1.2) u ψ C ψ L ψ C1 (), in which case we set [u] X = sup u ψ. ψ C 1 () ψ L 1 Note that if u X, then there exists a unique T [ C() ] = M() such that T, ψ = u ψ ψ C 1 (). On the other hand, by the Riesz Representation Theorem any T M() admits a unique decomposition T, ψ = ψ dν + ψ dµ ψ C(), where µ M() and ν M( ). As usual, M() and M( ) denote the spaces of finite measures in and, respectively, equipped with the norm M ; measures in M() are identified with measures in which do not charge. When u X, we will denote µ = u and ν =. Throughout the paper, whenever u X we use the notation u and in the above sense. If u X, then u ψ = ψ ψ u ψ C 1 (), and consequently, u ψ u ψ = Also, note that if u X, then [u] X = ψ ψ u u +. ψ C 2 (). In particular, [ ] X defines a seminorm in X and [u] X = 0 if, and only if, u is constant in. In order to verify this last assertion, one may use the fact that for every h C () with h = 0, there exists ψ C () such that ψ = h in with ψ = 0 on. Clearly, any function u W 2,1 () belongs to X and our notation is consistent with the usual meaning of u and. Recall that, for any function u L1 (), u is well-defined as a distribution. When u X, the distribution u belongs to M(), but the converse is not true; see, e.g., Proposition A.1 below. We now present our main results.
3 KATO S INEQUALITY UP TO THE BOUNDARY 3 Theorem 1.1. If u X, then u + X and (1.3) [u + ] X [u] X. In other words, (1.4) u u +. Our next result gives additional properties when u vanishes on the boundary: Theorem 1.2. If u W 1,1 0 () and u M() (in the sense of distributions), then u X (hence u + X). Moreover, (1.5) u + u. In addition, L1 ( ) with (1.6) u. Note that assertions (1.5) (1.6) fail if u does not vanish on ; simply take = B 1, the unit ball in R N, and u(x) = x 1. We now state our extension of Kato s inequality up to the boundary: Theorem 1.3. Let u X be such that u L 1 () and L1 ( ). Then, (1.7) u + ψ Hψ Gψ ψ C 1 (), ψ 0 in, where G L 1 () and H L 1 ( ) are given by { u on [u > 0], on [u > 0], (1.8) G = and H = 0 on [u < 0], 0 on [u 0], min {, 0} on [u = 0]. Thus, u + G in, (1.9) + H on. We conclude this introduction with the following problems: Open Problem 1. Let u X. Is it true that (1.10) + on? This problem is open even under the additional assumption that u W 1,1 0 (). Open Problem 2. Assume that u X and L1 ( ). Is it true that + L 1 ( )? More precisely, does one have + (1.11) = H, where H is the function given by (1.8)?
4 4 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) The answer to both Open Problems 1 and 2 is positive if u W 2,1 (); see Theorem 7.1 below. Addendum. Recently, A. Ancona informed us that he gave a positive answer to Open Problems 1 and 2 in full generality. His argument strongly relies on tools from Potential Theory; see [2]. 2. Properties of functions in X In this section, we investigate properties satisfied by elements in X. We first show that condition (1.2) required for a function to belong to X can be replaced by (2.1) u ζ C ζ L ζ C2 N(), where (2.2) C 2 N() = { ζ C 2 (); } ζ = 0 on. Proposition 2.1. Let u L 1 (). Then, u X if, and only if, (2.3) sup u ζ <. ζ C 2 N () ζ L 1 Moreover, (i) the quantity in (2.3) equals [u] X ; (ii) u W 1,p () for every 1 p < N N 1 ; moreover, u L p () C[u] X. In the proof of Proposition 2.1, we need the following variant of the classical De Giorgi-Stampacchia estimate (see [7, 8]) for the Neumann problem: Lemma 2.1. Given F C0 (; R N ), let w be the unique solution of w = div F in, (2.4) w = 0 on, such that w = 0. Then, for every q > N we have (2.5) w L C F L q. We present a sketch of the proof of Lemma 2.1 in Appendix C. Proof of Proposition 2.1. Note that if u X, then (2.6) u ζ = u ζ [u] X ζ L ζ C 2 N(). This gives the implication. We now assume that (2.3) holds. We split the proof of the converse into two steps: Step 1. u W 1,p () for every 1 p < where K denotes the quantity in (2.3). N N 1 and u L p () CK,
5 KATO S INEQUALITY UP TO THE BOUNDARY 5 Clearly, we may assume that 1 < p < N 1. Given F C 0 (; R N ), let w be the unique solution of (2.4) such that w = 0. By (2.3) and (2.5), we have u div F = u w K w L KC F L F p C 0 (; R N ). The conclusion follows by duality. Step 2. u X and [u] X = K. It suffices to show that (2.7) u ψ K ψ L N ψ C1 (). Indeed, this implies u X and [u] X K. Since by (2.6), K [u] X, equality must hold. We now turn ourselves to the proof of (2.7). Given ψ C 2 (), we first show that there exists a sequence (ζ k ) such that (2.8) ζ k C 2 N(), ζ k L C, ζ k ψ uniformly in and (2.9) ζ k ψ a.e. in. Indeed, let Φ C 0 (R) and η C 2 () with η = 0 on be such that Take Φ(t) = t t [ 1, 1] and η = ψ on. ζ k = ψ 1 Φ(kη) in. k Clearly, (2.8) holds. On the other hand, [ ] 1 k Φ(kη) = Φ (kη) η χ [η=0] η in. Since η = 0 a.e. on the set [η = 0], (2.9) follows. For every k 1, we thus have u ζ k = u ζ k K ζ k L. As k, we obtain (2.7) with test functions ψ C 2 (). argument, one then gets (2.7). The proof is complete. Using a density Remark 2.1. Using Proposition 2.1, one deduces that given measures µ M() and ν M( ), the Neumann problem u = µ in, (2.10) = ν on, has a solution u X if, and only if, (2.11) µ() + ν( ) = 0. The solution is unique up to an additive constant and belongs to u W 1,p () for every 1 p < N N 1. In particular, if u = 0, then u W 1,p () C[u] X. The following result complements Proposition 2.1:
6 6 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) Proposition 2.2. Let u L 1 () be such that (2.12) u ζ ζ dν + ζ dµ ζ CN(), 2 ζ 0 in for some µ M() and ν M( ). Then, u X, ) (2.13) [u] X 2 ( µ + M() + ν + M( ) and (2.14) Proof. By (2.12), we have (2.15) u ζ u µ in, ν on. ζ dν + + ζ dµ + ζ CN(), 2 ζ 0 in. For every ζ CN 2 (), we apply (2.15) with test functions ζ L ± ζ to get ( (2.16) u ζ 2 µ + M() + ν + M( ) ) ζ L. By Proposition 2.1, it follows that u X and (2.13) holds. Proceeding as in Step 2 of the proof of Proposition 2.1 (more precisely, using (2.8) (2.9)), one deduces from (2.12) that Therefore, This gives (2.14). u ψ ψ dν + ψ dµ ψ C 2 (), ψ 0 in. ψ ψ u ψ dν + ψ dµ ψ C 2 (), ψ 0 in. 3. Proof of Theorem 1.1 We begin by establishing the following lemma: Lemma 3.1. If u C 2 (), then (3.1) u + ψ ψ ψ u ψ C 1 (), ψ 0 in. Proof. We first prove the Claim. If u C 2 () and Φ C 2 (R) is convex, then (3.2) Φ(u) ψ ψ Φ (u) ψ Φ (u) u ψ C 1 (), ψ 0 in. Note that and, by the convexity of Φ, Φ(u) = Φ (u) on Φ(u) Φ (u) u in.
7 KATO S INEQUALITY UP TO THE BOUNDARY 7 Thus, for every ψ C 1 (), ψ 0 in, Φ(u) ψ = ψ Φ(u) ψ Φ(u) This establishes the claim. ψ Φ (u) ψ Φ (u) u. We now apply (3.2) with Φ = Φ k, where (Φ k ) is a sequence of smooth convex functions such that Φ k (0) = 0, Φ k L 1 and satisfying { Φ 1 if t 0, k(t) 0 if t < 0. As k, we obtain (3.1). We now prove a special case of Theorem 1.1 for functions in C 2 (): Lemma 3.2. Let u C 2 (). Then, u + X and (3.3) [u + ] X [u] X. Proof. Note that u + W 1,1 (). In order to establish the lemma, it thus suffices to show that (3.4) u + ψ [u] X ψ L ψ C 1 (). For this purpose, given ψ C 1 () we apply (3.1) with ψ = ψ L + ψ. We then get ( ) (3.5) u + ψ u ψ L + ψ ψ u. Since estimate (3.5) becomes ( u + ψ ( [u<0] u = [u<0] [u<0] + [u<0] ) u ψ L + u, ψ ) + u ψ L = [u] X ψ L. ψ u This relation holds for every ψ C 1 (). Replacing ψ by ψ, we obtain (3.4). This establishes the lemma. Proof of Theorem 1.1. Since u X, u ψ = ψ ψ u Taking ψ = 1 as a test function, we get (3.6) = u. ψ C 1 ().
8 8 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) Let (µ k ) C () and (ν k ) C ( ) be two sequences such that µ k u weak in M() and µ k L 1 () u M(), ν k weak in M( ) and ν k L 1 ( ). M( ) In view of (3.6) we may also assume that ν k = µ k k 1. For each k 1, let u k C 2 () be the unique function such that u k = µ k in, k = ν k on, and u k = u. Then, by Remark 2.1 applied to u k u, the sequence (u k) is bounded in W 1,p () for every 1 p < N N 1. Since u k u a.e., one deduces that u + k u+ weakly in L 1 (). On the other hand, applying Lemma 3.2 to u k, we get u + k ψ [u + k ] X ψ L [u k ] X ψ L ψ C 1 (). As k, we obtain from which the conclusion follows. u + ψ [u] X ψ L 4. Properties of ψ C 1 (), We start with a result which seems intuitively true, but still requires a proof: Proposition 4.1. Let u W 1, (). Then, u X if, and only if, u M() (in the sense of distributions). In this case, L ( ) and (4.1) u L (). L ( ) If u C 1 () X, then coincides with the standard normal derivative on. Proof. We first assume that u W 1, () and u M(). Given a sequence of mollifiers (ρ k ) such that supp ρ k B 1/k, let u k (x) = ρ k (x y)u(y) dy x. Note that if d(x, ) > 1/k, then u k (x) = ρ k (x y) u(y) dy and u k (x) = ρ k (x y) u(y) dy.
9 KATO S INEQUALITY UP TO THE BOUNDARY 9 Denote (4.2) δ = { x ; d(x, ) > δ } ; for δ 0 > 0 small enough, δ is smooth for every δ (0, δ 0 ). For every k 1 and δ (0, δ 0 ) such that 1/k < δ we then have (4.3) k u k L ( δ ) u L (). L ( δ ) Thus, for every ψ C 1 (), (4.4) ψ u k + ψ u k u L () ψ L 1 ( δ ). δ δ Note that for a.e. δ (0, δ 0 ) (4.5) δ u = 0; hence, for any such δ > 0, ψ u k δ ψ u δ as k. Indeed, this is a general fact (see, e.g., [5, Theorem 1, p.54]): if µ M() and µ ( δ ) = 0, then ψ (ρ k µ) ψ dµ ψ C 0 ( δ ). δ δ For any δ (0, δ 0 ) verifying (4.5), as k in (4.4) we get (4.6) ψ u + ψ u u L () ψ L 1 ( δ ) δ δ ψ C 1 (). From this estimate, one deduces that for every ψ C 1 (), ψ u u M() ψ L ( δ ) + u L () ψ L 1 ( δ ) δ As δ 0, we conclude that u X. ( u M() + u L () δ ) ψ L (). In order to prove that L ( ), we return to estimate (4.6). Given φ C 1 ( ), we fix an extension ψ C 1 () of φ; note that ψ L 1 ( δ ) φ L 1 ( ) + Cδ δ (0, δ 0 ), for some constant C > 0. Insert this test function ψ in (4.6). As δ 0 we obtain, by dominated convergence, ψ u + ψ u u L () φ L 1 ( ). Hence, φ u L () φ L1 ( ) Therefore, by duality L ( ) and (4.1) holds. φ C 1 ( ).
10 10 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) We now assume that u C 1 () X and we denote by h the normal derivative of u in the standard sense. By Lemma B.1 and Remark B.1, there exists a sequence (u k ) C () satisfying (B.2) (B.3) and such that In particular, Thus, (4.7) u k u k h Hence, the normal derivative in C 1 (). uniformly on. u ψ + ψ u = h ψ ψ C 1 (). in the sense of the space X coincides with h. When u X the measure need not be an L1 -function. Surprisingly, this is always true if u vanishes on : Proposition 4.2. Let u W 1,1 0 (). Then, u X if, and only if, u M() in the sense of distributions. Moreover, L1 ( ) and (4.8) u M(). L 1 ( ) Proof. We split the proof into two steps: Step 1. Proof of (4.8) if u is smooth in a neighborhood of. Under this assumption, is a smooth function on. Denote by v 1 and v 2 the solutions of { v1 = µ + in, { v2 = µ in, v 1 = 0 where µ = u. In particular, on, u = v 1 v 2 in. v 2 = 0 on, Since µ is smooth in a neighborhood of, µ + and µ are Lipschitz continuous near. Hence, v 1 and v 2 are of class C 2 near. Moreover, v 1 0 in and v 1 = 0 on ; thus, It follows that Similarly, Therefore, v 1 v 1 0 = v 1 + v 2 = Step 2. Proof of the proposition completed. on. v 1 = µ +. µ. v 2 = (µ + + µ ) = u.
11 KATO S INEQUALITY UP TO THE BOUNDARY 11 Let (ϕ k ) C 0 () be a sequence of test functions such that 0 ϕ k 1 in and ϕ k (x) = 1 if d(x, ) 1 k. Take µ k = ϕ k u, k 1. Then, (µ k ) M() is a sequence of measures such that supp µ k and, by dominated convergence, (4.9) µ k u strongly in M(). For each k 1, let u k be the unique solution of { uk = µ k in, u k = 0 on. Note that u k is harmonic in a neighborhood of. We claim that (4.10) φ k φ φ C1 ( ). Indeed, since u k u in L 1 () and ( u k ) is bounded in W 1,p 0 () for every 1 p < N N 1, (see [10]) we have (4.11) ψ u k ψ u ψ C 1 (). Assertion (4.10) then follows from (4.9) and (4.11). Applying Step 1 to the function u i u j, we have i j µ i µ j M() i, j 1. L1 ( ) In view of the strong convergence of (µ k ) in M(), ( k ) is a Cauchy sequence in L 1 ( ). Hence, this sequence converges in L 1 ( ) to some function h. By (4.10), h = ; hence, k in L1 ( ). Moreover, since (4.8) holds for every u k, it also holds for u. The proof is complete. We now show that if u W 1,1 () and u BV () then the normal derivative in the sense of the space X coincides with the function n u on defined in the sense of traces: Proposition 4.3. Assume that u W 1,1 () and u BV (); hence, u = div ( u) M(). Then, u X and coincides with n u on, where u is understood in the sense of traces. In particular, L1 ( ) and (4.12) C u BV (). L 1 ( ) In the proof of Proposition 4.3 we use the notion of strict convergence in BV (A), where A R N is a Lipschitz domain. We recall that a sequence (f n ) BV (A) converges strictly to f BV (A) if f n f strongly in L 1 (A) and Df n Df. A A
12 12 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) By [1, Theorem 3.88], the trace operator f BV (A) f A L 1 ( A) is continuous from BV (A) (under strict convergence) into L 1 ( A) (under strong convergence). Proof of Proposition 4.3. By Lemma B.1 and Remark B.1, there exists a sequence (u k ) C () satisfying (B.1) (B.3) and (B.12). Since ( u k ) converges strictly to u in BV ( ), we have (4.13) u k u in L 1 ( ). Hence, ( ) u ψ + ψ u = n u ψ ψ C 1 ( ). This implies that L1 ( ) and equals n u. By the BV -trace theory, (4.12) holds. 5. Proof of Theorem 1.2 We first establish Theorem 1.2 for functions in CD 2 (), where { } (5.1) CD() 2 = ζ C 2 (); ζ = 0 on. Lemma 5.1. Let u C 2 D (). Then, u+ M() and (5.2) u + M u L 1. Proof. Apply (3.3) with u + a, where a > 0. We deduce that [ (5.3) (u + a) + ] [u + a] X X = [u] X. Since (u + a) + = u + a in a neighborhood of, (5.4) Note that (u + a)+ = on. [ (u + a) + ] = X (u + a) + M() + (u + a)+ [u] X = u L1 () +. L1 ( ) By (5.3) (5.4) we then have (u + a) + M u L 1 a > 0. L 1 ( ) The result follows from the lower semicontinuity of the norm M with respect to the weak convergence as a 0. Proof of Theorem 1.2. Since u X, u M(). Take a sequence (µ k ) C () such that µ k u weak in M() and µ k L 1 µ M. For each k 1, let u k CD 2 () be the solution of u k = µ k in.,
13 KATO S INEQUALITY UP TO THE BOUNDARY 13 Then, by standard elliptic estimates, u k u in L 1 (). On the other hand, it follows from Lemma 5.1 that u + k Thus, As k we obtain u + k M u k L 1. M() and u + k ζ u k L 1 ζ L = µ k L 1 ζ L ζ C 2 D(). u + ζ u M ζ L ζ C 2 D(). This gives (1.5). From Proposition 4.2, we know that holds., + 6. Kato s inequality up to the boundary L1 ( ) and (1.6) Before proving Theorem 1.3, we first present some variants of Kato s inequality when u and are not necessarily L1 -functions but only finite measures. We prove for instance the following companion to [3, Proposition 4.B.5]: Proposition 6.1. Let u L 1 () be such that (6.1) u ζ hζ + gζ ζ CN(), 2 ζ 0 in for some g L 1 () and h L 1 ( ). Then, u W 1,1 () and (6.2) u + ψ hψ + gψ ψ C 1 (), ψ 0 in. Proof. By Proposition 2.2, u X. Moreover, u g in, (6.3) h on. We now split the proof into two steps: Step 1. Let Φ C 2 (R) be a nondecreasing convex function such that Φ L (R). Then, (6.4) Φ(u) ψ ψ Φ (u)h + ψ Φ (u)g for every ψ C 1 () such that ψ 0 in. Let (g k ) C () and (h k ) C ( ) be such that g k g in L 1 () and a.e. and h k h in L 1 ( ) and a.e. Next, take (µ k ) C () and (ν k ) C ( ) such that µ k u weak in M() and ν k weak in M( ).
14 14 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) In view of (6.3) and we may assume that = u, µ k g k in, ν k h k on and ν k = µ k k 1. Let u k C () be the unique solution of u k = µ k in, k = ν k on, such that u k = u. By Remark 2.1, the sequence (u k) is bounded in W 1,p () for every 1 p <. Passing to a subsequence if necessary, we have N N 1 Φ(u k ) Φ(u) weakly in L 1 (). Let ψ C 1 (), ψ 0 in. As in Lemma 3.1, for every k 1 we have Φ(u k ) ψ ψ Φ (u k ) k ψ Φ (u k ) u k ψ Φ (u k )h k + ψ Φ (u k )g k. By dominated convergence we obtain (6.4) as k. Step 2. Proof of the proposition completed. Apply (6.4) with Φ = Φ k, where (Φ k ) is a sequence of smooth convex functions such that Φ k (0) = 0, 0 Φ k 1 and { Φ 1 if t 0, k(t) 0 if t < 0. The result follows as we let k. The following variant of Proposition 6.1 will be needed below: Proposition 6.2. Let u L 1 () be such that (6.5) u ζ hζ + ζ dµ ζ CN(), 2 ζ 0 in for some µ M(), µ 0, and h L 1 ( ). Then, u W 1,1 () and (6.6) u + ψ hψ + ψ dµ ψ C 1 (), ψ 0 in. Proof. One can proceed as in the proof of Proposition 6.1. In Step 1, one should replace (6.4) by (6.4 ) Φ(u) ψ ψ Φ (u)h + Φ L ψ dµ. Inequality (6.4 ) is easily obtained by approximation, where the sequence (g k ) C () is chosen so that g k µ weak in M().
15 KATO S INEQUALITY UP TO THE BOUNDARY 15 The rest of the argument remains unchanged. We now prove the Proposition 6.3. Let u X. If L1 ( ), then + on [u > 0], (6.7) 0 on [u < 0], min {, 0} on [u = 0]. Proof. Denoting by µ = ( u) + and h =, we have u ζ hζ + ζ dµ ζ CN(), 2 ζ 0 in. Therefore, by Proposition 6.2, u + satisfies (6.8) u + ψ hψ + ψ dµ ψ C 1 (), ψ 0 in. By Theorem 1.1, we know that u + X. It thus follows that + (6.9) χ h = χ on. Given a > 0, we now apply (6.8) with u replaced by u a. As a 0, we obtain (6.10) u + ψ u + ψ hψ + ψ dµ ψ C 1 (), ψ 0 in. Hence, In particular, [u>0] + χ [u>0]h = χ [u>0] + (6.11) 0 on [u = 0]. Assertion (6.7) follows by combining (6.9) and (6.11). on. We state the following consequence of Proposition 6.3: Corollary 6.1. Let u X W 1,1 0 (). If u 0 in, then 0 on. Proof. Since u = u + in and u = 0 on, applying Proposition 6.3 above we get { } = + min, 0 0 on. We now present the Proof of Theorem 1.3. By Theorem 1.1, u + X. Applying Kato s inequality to u a, we have (6.12) (u a) + χ [u a] u in for every a R. As a 0 in (6.12) we get u + χ [u>0] u = G in.
16 16 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) By this estimate and (6.7), for every ψ C 1 () with ψ 0 in, u + ψ = ψ + ψ u + Hψ The proof is complete. Gψ. 7. Computing + for W 2,1 -functions Our goal in this section is to give a positive answer to Open Problems 1 and 2 under the additional assumption that u W 2,1 (): Theorem 7.1. If u W 2,1 (), then u + BV () (so that, u + X by Proposition 4.3) and + on [u > 0], (7.1) = 0 on [u < 0], min {, 0} on [u = 0]. We first prove the Lemma 7.1. If v W 1,1 () and v BV (), then v v(x) v ( x tn(x) ) (7.2) (x) = lim H N 1 -a.e. on. t 0 t In (7.2), we identify v with its precise representative, which is well-defined outside a set of zero H N 1 -Hausdorff measure; see [5, Section 4.8, Theorem 1 and Section 5.6, Theorem 3]. Proof. Since v W 1,1 (), for H N 1 -a.e. x the function t (0, δ) v ( x tn(x) ) is well-defined for some δ > 0 and belongs to W 1,1 (0, δ). Thus, (7.3) v ( x tn(x) ) v(x) t = n(x) 1 0 v ( x stn(x) ) ds. Moreover, since v BV (), for H N 1 -a.e. x the function r (0, δ) v ( x rn(x) ) belongs to BV (0, δ) L (0, δ) and (see [1, Theorem 3.108]) (7.4) lim r 0 v ( x rn(x) ) = v (x). We deduce from (7.3) (7.4) that ( ) v x tn(x) v(x) lim = n(x) v (x). t 0 t By Proposition 4.3 above, v = n v and the conclusion follows. We also need the following elementary lemma whose proof is left to the reader:
17 KATO S INEQUALITY UP TO THE BOUNDARY 17 Lemma 7.2. Let v : [0, δ] R be such that v(0) v(t) (7.5) lim = α R. t 0 t Then, v + (0) v + (t) (7.6) lim = t 0 t We now present the α if v(0) > 0, 0 if v(0) < 0, min {α, 0} if v(0) = 0. Proof of Theorem 7.1. We split the proof into three steps: Step 1. Proof of the assertion: u + BV (). (7.7) Extending u to R N, we may assume that u W 2,1 (R N ). We claim that 2 u + e 2 χ 2 u e 2 in D (R N ) for every e R N \ {0}. Indeed, let (Φ k ) be a sequence of smooth convex functions such that Φ k (0) = 0, Φ k L 1 and { (7.8) Φ 1 if t 0, k(t) 0 if t < 0. Then, 2 Φ k (u) e 2 ( ) 2 = Φ k(u) 2 u e 2 + Φ k(u) Φ e k(u) 2 u e 2 in R N. As k, we obtain (7.7). It follows from (7.7) that 2 u + e 2 M() for every e R N \ {0}. Applying the conclusion with e = e i, e j, e i + e j for every i, j {1,..., N} we deduce that D 2 u + is a finite measure in. Thus, u + BV (). Step 2. Proof of (7.1). By Lemma 7.1, for H N 1 -a.e. x, u satisfies ( ) u(x) u x tn(x) (7.9) lim = t 0 t (x). Hence, by (7.2) applied to u + and by (7.6) applied to v(t) = u(x tn(x)), + u + (x) u +( x tn(x) ) (x) if u(x) > 0, (x) = lim = 0 if u(x) < 0, t 0 t min { (x), 0} if u(x) = 0, for every x for which (7.9) holds. Since this is true H N 1 -a.e. on, (7.1) follows. The proof of Theorem 7.1 is complete.
18 18 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) Appendix A. The measure u + need not be finite In this appendix, we construct a harmonic function in dimension 2 such that u+ = : Proposition A.1. Let = { (x, y) R 2 ; x 2 + y 2 < 1 and x > 0 }. There exists a harmonic function u C() H 1 () with u W 1,1 ( ) such that (i) u X and u + X; (ii) u + 0 in the sense of distributions; (iii) u + is not a finite measure in. Proof. Let u be the function in given in polar coordinates by (A.1) u(r, θ) = r a k sin (a k θ) k=1 where (a k ) (0, 1) is a sequence such that ka k <. Since u(r, θ) k=1 sin (a k θ) π 2 k=1 a k, it follows that u C() and u is harmonic in (u is a series of harmonic functions). Note that u 2 = a j a k r aj+ak 2 cos ( (a j a k )θ ). Thus, u 2 π j,k=1 j,k=1 a j a k a j + a k 2π j,k=1 j k k=1 a j a k a j + a k 2π ka k < ; in other words, u H 1 (). Denoting by τ the tangential unit vector of u on, we have τ = 4 ( π ) sin a k 2π a k < ; 2 k=1 hence, u W 1,1 ( ). Since u is harmonic in, u + is subharmonic. Thus, u + 0 in. We show that u + is not a finite measure in. Note that u vanishes only on the x-axis. Denoting by dx (= dr) the 1-dimensional Lebesgue measure on the segment (0, 1) {0}, we then have u + = y (x, 0) dx = 1 r θ (r, 0) dr = a k r ak 1 dr. Therefore, u + = k=1 1 0 a k r a k 1 dr = k=1 k=1 k=1 1 =. k=1
19 KATO S INEQUALITY UP TO THE BOUNDARY 19 Hence, u + X and, by Theorem 1.1, this means that u X. Remark A.1. This example also shows that given ϕ W 1,1 ( ), it is in general not possible to construct a function v W 2,1 () such that v = ϕ. This is in contrast with the well-known result of Gagliardo [6] which asserts that the map w W 1,1 () w L 1 ( ) is surjective. Indeed, take ϕ = u, where u is given by (A.1). Suppose by contradiction that there exists some v W 2,1 () such that v = ϕ. Applying Proposition 4.2 to u v W 1,1 0 (), we would deduce that (u v) L1 ( ). But v W 2,1 () implies v L1 ( ) and therefore = v (u v) + L1 ( ), a contradiction. Appendix B. Approximation by smooth functions in In this appendix, we establish the following Lemma B.1. Given u X, there exists a sequence (u k ) C () such that (B.1) (B.2) and (B.3) u k u in W 1,1( ), ψ u k ψ u ψ C 1 () ψ k ψ ψ C1 (). Proof. We split the proof into two steps: Step 1. Given x 0, there exist δ > 0 and a sequence (v k ) C () such that (B.4) (B.5) v k u in W 1,1( B δ (x 0 ) ), ψ v k ψ u ψ C 1 () with supp ψ B δ (x 0 ). Since is smooth, there exist δ 1 > 0 and an open cone T R N (with vertex at 0 R N ) such that (B.6) (x + T ) B δ1 (x) x B δ1 (x 0 ). Let δ = δ 1 /2 and ρ C 0 (B δ ), ρ 0, be such that B δ ρ = 1 and (B.7) Set supp ρ T. ρ k (x) = k N ρ(kx) x R N. We show that the sequence (v k ) C () given by (B.8) v k (x) = ρ k (x y)u(y) dy x satisfies (B.4) (B.5). Note that given any x B δ (x 0 ), by (B.7) v k (x) depends only on the values of
20 20 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) u on a compact subset of (x + T ) B δ1 (x). In fact, from (B.6) (B.7) and a change of variable, we can rewrite (B.8) as (B.9) v k (x) = ρ k ( z)u(x + z) dz x B δ (x 0 ). Therefore, T B δ1 (0) (B.10) v k = ρ k ( u) and v k = ρ k ( u) in B δ (x 0 ). In particular, (B.4) (B.5) hold. Step 2. Proof of the proposition completed. By compactness of, we can cover this set with finitely many balls B δ (x 1 ),..., B δ (x t ) such that (B.4) (B.5) hold on each ball B δ (x i ) for some sequence (v i k ) C (). We now take (v 0 k ) C () and ω such that \ t i=1 B δ(x i ) ω, v 0 k u in W 1,1 (ω) and v 0 k u weak in M(ω) (such sequence can be obtained via convolution of u). Let (ϕ i ) be a partition of unity subordinated to the covering ω, B δ (x 1 ),..., B δ (x t ) of. One verifies that (B.1) (B.2) hold for the sequence (u k ) given by u k = t ϕ i vk. i i=0 Assertion (B.3) immediately follows from (B.1) (B.2). Remark B.1. An inspection of the proof of Lemma B.1 shows that (i) if u C 1 (), then (B.11) u k u in C 1 (); (ii) if u BV (), then (B.12) D 2 u k L 1 () D 2 u M(). Appendix C. Proof of Lemma 2.1 The proof of Lemma 2.1 we present below follows the lines of [8, Lemma 7.3] (see also [7, Theorem 8.15]) with some minor modifications. We first need the following variant of the Gagliardo-Nirenberg inequality: Proposition C.1. Let (C.1) A = Then, { v W 1,1 (); } [v = 0]. 3 (C.2) v L N N 1 C v L 1 v A. We denote by E the Lebesgue measure of a set E R N.
21 KATO S INEQUALITY UP TO THE BOUNDARY 21 Proof. By a variant of the Poincaré inequality (easily proved by contradiction), we have (C.3) v L 1 C v L 1 v A. On the other hand, by the standard Gagliardo-Nirenberg inequality and an extension argument, (C.4) v N L N 1 C ( ) v L 1 + v L 1 v W 1,1 (). Combining (C.3) (C.4), we obtain (C.2). Proof of Lemma 2.1. Replacing w by w a for some suitable constant a R if necessary, we may assume that (C.5) Given t > 0, let [w 0] 3 and [w 0] 3. (C.6) v t (x) = [ w(x) t ] + x. Using v t as a test function in (2.4), one shows that v t L 2 1 F L q [w > t] 2 1 q. On the other hand, by Hölder s inequality and Proposition C.1, 1 v t L 1 C v t L 2 [w > t] N. Thus, (C.7) v t L 1 C F L q [w > t] α t > 0, where α = N 1 q. Recall that (C.8) v t L 1 = 0 [vt > r] M dr = [w > s] ds, where M = w + L. Since α > 1, one deduces using (C.7) (C.8) that (C.9) w + L C F 1 α L q w α L 1. From (C.9) and w + L 1 w + L, we then have w + L C F L q. Replacing w by w, one obtains a similar estimate for w. Thus, t w L w + L + w L 2C F L q. Acknowledgment. The work of the first author (H. B.) is partially supported by NSF Grant DMS
22 22 HAÏM BREZIS(1),(2) AND AUGUSTO C. PONCE (3) References [1] L. Ambrosio, N. Fusco, and D. Pallara, Functions of bounded variation and free discontinuity problems. Oxford Mathematical Monographs, Oxford University Press, New York, [2] A. Ancona, Inégalité de Kato et inégalité de Kato jusqu au bord. C. R. Math. Acad. Sci. Paris 346 (2008), [3] H. Brezis, M. Marcus, and A. C. Ponce, Nonlinear elliptic equations with measures revisited. In: Mathematical Aspects of Nonlinear Dispersive Equations (J. Bourgain, C. Kenig, and S. Klainerman, eds.), Annals of Mathematics Studies, 163, Princeton University Press, Princeton, NJ, 2007, pp [4] H. Brezis and A. C. Ponce, Kato s inequality when u is a measure. C. R. Math. Acad. Sci. Paris, Ser. I 338 (2004), [5] L. C. Evans and R. F. Gariepy, Measure theory and fine properties of functions. Studies in Advanced Mathematics, CRC Press, Boca Raton, FL, [6] E. Gagliardo, Caratterizzazioni delle tracce sulla frontiera relative ad alcune classi di funzioni in n variabili. Rend. Sem. Mat. Univ. Padova 27 (1957), [7] D. Gilbarg and N. S. Trudinger, Elliptic partial differential equations of second order. Grundlehren der Mathematischen Wissenschaften, vol. 224, Springer-Verlag, Berlin, [8] P. Hartman and G. Stampacchia, On some non-linear elliptic differential-functional equations. Acta Math. 115 (1966), [9] T. Kato, Schrödinger operators with singular potentials. Israel J. Math. 13 (1972), (1973). Proceedings of the International Symposium on Partial Differential Equations and the Geometry of Normed Linear Spaces (Jerusalem, 1972). [10] G. Stampacchia, Équations elliptiques du second ordre à coefficients discontinus. Séminaire de Mathématiques Supérieures, no. 16 (été, 1965), Les Presses de l Université de Montréal, Montréal, Québec, (1) Rutgers University Department of Mathematics, Hill Center, Busch Campus 110 Frelinghuysen Rd. Piscataway, NJ 08854, USA (2) Technion Department of Mathematics Haifa, Israel (3) Université catholique de Louvain Département de mathématique Chemin du Cyclotron Louvain-la-Neuve, Belgium
No: 10 04. Bilkent University. Monotonic Extension. Farhad Husseinov. Discussion Papers. Department of Economics
No: 10 04 Bilkent University Monotonic Extension Farhad Husseinov Discussion Papers Department of Economics The Discussion Papers of the Department of Economics are intended to make the initial results
More informationDuality of linear conic problems
Duality of linear conic problems Alexander Shapiro and Arkadi Nemirovski Abstract It is well known that the optimal values of a linear programming problem and its dual are equal to each other if at least
More informationMATH 425, PRACTICE FINAL EXAM SOLUTIONS.
MATH 45, PRACTICE FINAL EXAM SOLUTIONS. Exercise. a Is the operator L defined on smooth functions of x, y by L u := u xx + cosu linear? b Does the answer change if we replace the operator L by the operator
More informationINDISTINGUISHABILITY OF ABSOLUTELY CONTINUOUS AND SINGULAR DISTRIBUTIONS
INDISTINGUISHABILITY OF ABSOLUTELY CONTINUOUS AND SINGULAR DISTRIBUTIONS STEVEN P. LALLEY AND ANDREW NOBEL Abstract. It is shown that there are no consistent decision rules for the hypothesis testing problem
More information0 <β 1 let u(x) u(y) kuk u := sup u(x) and [u] β := sup
456 BRUCE K. DRIVER 24. Hölder Spaces Notation 24.1. Let Ω be an open subset of R d,bc(ω) and BC( Ω) be the bounded continuous functions on Ω and Ω respectively. By identifying f BC( Ω) with f Ω BC(Ω),
More informationSeparation Properties for Locally Convex Cones
Journal of Convex Analysis Volume 9 (2002), No. 1, 301 307 Separation Properties for Locally Convex Cones Walter Roth Department of Mathematics, Universiti Brunei Darussalam, Gadong BE1410, Brunei Darussalam
More informationGeometrical Characterization of RN-operators between Locally Convex Vector Spaces
Geometrical Characterization of RN-operators between Locally Convex Vector Spaces OLEG REINOV St. Petersburg State University Dept. of Mathematics and Mechanics Universitetskii pr. 28, 198504 St, Petersburg
More informationON NONNEGATIVE SOLUTIONS OF NONLINEAR TWO-POINT BOUNDARY VALUE PROBLEMS FOR TWO-DIMENSIONAL DIFFERENTIAL SYSTEMS WITH ADVANCED ARGUMENTS
ON NONNEGATIVE SOLUTIONS OF NONLINEAR TWO-POINT BOUNDARY VALUE PROBLEMS FOR TWO-DIMENSIONAL DIFFERENTIAL SYSTEMS WITH ADVANCED ARGUMENTS I. KIGURADZE AND N. PARTSVANIA A. Razmadze Mathematical Institute
More informationFuzzy Differential Systems and the New Concept of Stability
Nonlinear Dynamics and Systems Theory, 1(2) (2001) 111 119 Fuzzy Differential Systems and the New Concept of Stability V. Lakshmikantham 1 and S. Leela 2 1 Department of Mathematical Sciences, Florida
More informationNotes on metric spaces
Notes on metric spaces 1 Introduction The purpose of these notes is to quickly review some of the basic concepts from Real Analysis, Metric Spaces and some related results that will be used in this course.
More informationEXISTENCE AND NON-EXISTENCE RESULTS FOR A NONLINEAR HEAT EQUATION
Sixth Mississippi State Conference on Differential Equations and Computational Simulations, Electronic Journal of Differential Equations, Conference 5 (7), pp. 5 65. ISSN: 7-669. UL: http://ejde.math.txstate.edu
More informationMetric Spaces. Chapter 7. 7.1. Metrics
Chapter 7 Metric Spaces A metric space is a set X that has a notion of the distance d(x, y) between every pair of points x, y X. The purpose of this chapter is to introduce metric spaces and give some
More informationContinued Fractions and the Euclidean Algorithm
Continued Fractions and the Euclidean Algorithm Lecture notes prepared for MATH 326, Spring 997 Department of Mathematics and Statistics University at Albany William F Hammond Table of Contents Introduction
More information1 Completeness of a Set of Eigenfunctions. Lecturer: Naoki Saito Scribe: Alexander Sheynis/Allen Xue. May 3, 2007. 1.1 The Neumann Boundary Condition
MAT 280: Laplacian Eigenfunctions: Theory, Applications, and Computations Lecture 11: Laplacian Eigenvalue Problems for General Domains III. Completeness of a Set of Eigenfunctions and the Justification
More informationBANACH AND HILBERT SPACE REVIEW
BANACH AND HILBET SPACE EVIEW CHISTOPHE HEIL These notes will briefly review some basic concepts related to the theory of Banach and Hilbert spaces. We are not trying to give a complete development, but
More information1 if 1 x 0 1 if 0 x 1
Chapter 3 Continuity In this chapter we begin by defining the fundamental notion of continuity for real valued functions of a single real variable. When trying to decide whether a given function is or
More information8.1 Examples, definitions, and basic properties
8 De Rham cohomology Last updated: May 21, 211. 8.1 Examples, definitions, and basic properties A k-form ω Ω k (M) is closed if dω =. It is exact if there is a (k 1)-form σ Ω k 1 (M) such that dσ = ω.
More informationA PRIORI ESTIMATES FOR SEMISTABLE SOLUTIONS OF SEMILINEAR ELLIPTIC EQUATIONS. In memory of Rou-Huai Wang
A PRIORI ESTIMATES FOR SEMISTABLE SOLUTIONS OF SEMILINEAR ELLIPTIC EQUATIONS XAVIER CABRÉ, MANEL SANCHÓN, AND JOEL SPRUCK In memory of Rou-Huai Wang 1. Introduction In this note we consider semistable
More informationMICROLOCAL ANALYSIS OF THE BOCHNER-MARTINELLI INTEGRAL
PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 00, Number 0, Pages 000 000 S 0002-9939(XX)0000-0 MICROLOCAL ANALYSIS OF THE BOCHNER-MARTINELLI INTEGRAL NIKOLAI TARKHANOV AND NIKOLAI VASILEVSKI
More informationUNIVERSITETET I OSLO
NIVERSITETET I OSLO Det matematisk-naturvitenskapelige fakultet Examination in: Trial exam Partial differential equations and Sobolev spaces I. Day of examination: November 18. 2009. Examination hours:
More informationDIRICHLET S PROBLEM WITH ENTIRE DATA POSED ON AN ELLIPSOIDAL CYLINDER. 1. Introduction
DIRICHLET S PROBLEM WITH ENTIRE DATA POSED ON AN ELLIPSOIDAL CYLINDER DMITRY KHAVINSON, ERIK LUNDBERG, HERMANN RENDER. Introduction A function u is said to be harmonic if u := n j= 2 u = 0. Given a domain
More informationTOPIC 4: DERIVATIVES
TOPIC 4: DERIVATIVES 1. The derivative of a function. Differentiation rules 1.1. The slope of a curve. The slope of a curve at a point P is a measure of the steepness of the curve. If Q is a point on the
More informationRESONANCES AND BALLS IN OBSTACLE SCATTERING WITH NEUMANN BOUNDARY CONDITIONS
RESONANCES AND BALLS IN OBSTACLE SCATTERING WITH NEUMANN BOUNDARY CONDITIONS T. J. CHRISTIANSEN Abstract. We consider scattering by an obstacle in R d, d 3 odd. We show that for the Neumann Laplacian if
More informationPacific Journal of Mathematics
Pacific Journal of Mathematics GLOBAL EXISTENCE AND DECREASING PROPERTY OF BOUNDARY VALUES OF SOLUTIONS TO PARABOLIC EQUATIONS WITH NONLOCAL BOUNDARY CONDITIONS Sangwon Seo Volume 193 No. 1 March 2000
More informationCONTINUOUS REINHARDT DOMAINS PROBLEMS ON PARTIAL JB*-TRIPLES
CONTINUOUS REINHARDT DOMAINS PROBLEMS ON PARTIAL JB*-TRIPLES László STACHÓ Bolyai Institute Szeged, Hungary stacho@math.u-szeged.hu www.math.u-szeged.hu/ Stacho 13/11/2008, Granada László STACHÓ () CONTINUOUS
More informationALMOST COMMON PRIORS 1. INTRODUCTION
ALMOST COMMON PRIORS ZIV HELLMAN ABSTRACT. What happens when priors are not common? We introduce a measure for how far a type space is from having a common prior, which we term prior distance. If a type
More informationn k=1 k=0 1/k! = e. Example 6.4. The series 1/k 2 converges in R. Indeed, if s n = n then k=1 1/k, then s 2n s n = 1 n + 1 +...
6 Series We call a normed space (X, ) a Banach space provided that every Cauchy sequence (x n ) in X converges. For example, R with the norm = is an example of Banach space. Now let (x n ) be a sequence
More informationFurther Study on Strong Lagrangian Duality Property for Invex Programs via Penalty Functions 1
Further Study on Strong Lagrangian Duality Property for Invex Programs via Penalty Functions 1 J. Zhang Institute of Applied Mathematics, Chongqing University of Posts and Telecommunications, Chongqing
More informationInner Product Spaces
Math 571 Inner Product Spaces 1. Preliminaries An inner product space is a vector space V along with a function, called an inner product which associates each pair of vectors u, v with a scalar u, v, and
More informationF. ABTAHI and M. ZARRIN. (Communicated by J. Goldstein)
Journal of Algerian Mathematical Society Vol. 1, pp. 1 6 1 CONCERNING THE l p -CONJECTURE FOR DISCRETE SEMIGROUPS F. ABTAHI and M. ZARRIN (Communicated by J. Goldstein) Abstract. For 2 < p
More informationMetric Spaces. Chapter 1
Chapter 1 Metric Spaces Many of the arguments you have seen in several variable calculus are almost identical to the corresponding arguments in one variable calculus, especially arguments concerning convergence
More informationHøgskolen i Narvik Sivilingeniørutdanningen STE6237 ELEMENTMETODER. Oppgaver
Høgskolen i Narvik Sivilingeniørutdanningen STE637 ELEMENTMETODER Oppgaver Klasse: 4.ID, 4.IT Ekstern Professor: Gregory A. Chechkin e-mail: chechkin@mech.math.msu.su Narvik 6 PART I Task. Consider two-point
More informationParabolic Equations. Chapter 5. Contents. 5.1.2 Well-Posed Initial-Boundary Value Problem. 5.1.3 Time Irreversibility of the Heat Equation
7 5.1 Definitions Properties Chapter 5 Parabolic Equations Note that we require the solution u(, t bounded in R n for all t. In particular we assume that the boundedness of the smooth function u at infinity
More informationx a x 2 (1 + x 2 ) n.
Limits and continuity Suppose that we have a function f : R R. Let a R. We say that f(x) tends to the limit l as x tends to a; lim f(x) = l ; x a if, given any real number ɛ > 0, there exists a real number
More informationClass Meeting # 1: Introduction to PDEs
MATH 18.152 COURSE NOTES - CLASS MEETING # 1 18.152 Introduction to PDEs, Fall 2011 Professor: Jared Speck Class Meeting # 1: Introduction to PDEs 1. What is a PDE? We will be studying functions u = u(x
More informationTHE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS
THE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS KEITH CONRAD 1. Introduction The Fundamental Theorem of Algebra says every nonconstant polynomial with complex coefficients can be factored into linear
More informationShape Optimization Problems over Classes of Convex Domains
Shape Optimization Problems over Classes of Convex Domains Giuseppe BUTTAZZO Dipartimento di Matematica Via Buonarroti, 2 56127 PISA ITALY e-mail: buttazzo@sab.sns.it Paolo GUASONI Scuola Normale Superiore
More informationMINIMIZATION OF ENTROPY FUNCTIONALS UNDER MOMENT CONSTRAINTS. denote the family of probability density functions g on X satisfying
MINIMIZATION OF ENTROPY FUNCTIONALS UNDER MOMENT CONSTRAINTS I. Csiszár (Budapest) Given a σ-finite measure space (X, X, µ) and a d-tuple ϕ = (ϕ 1,..., ϕ d ) of measurable functions on X, for a = (a 1,...,
More informationMEASURE AND INTEGRATION. Dietmar A. Salamon ETH Zürich
MEASURE AND INTEGRATION Dietmar A. Salamon ETH Zürich 12 May 2016 ii Preface This book is based on notes for the lecture course Measure and Integration held at ETH Zürich in the spring semester 2014. Prerequisites
More informationHOMEWORK 5 SOLUTIONS. n!f n (1) lim. ln x n! + xn x. 1 = G n 1 (x). (2) k + 1 n. (n 1)!
Math 7 Fall 205 HOMEWORK 5 SOLUTIONS Problem. 2008 B2 Let F 0 x = ln x. For n 0 and x > 0, let F n+ x = 0 F ntdt. Evaluate n!f n lim n ln n. By directly computing F n x for small n s, we obtain the following
More informationOptimal boundary control of quasilinear elliptic partial differential equations: theory and numerical analysis
Optimal boundary control of quasilinear elliptic partial differential equations: theory and numerical analysis vorgelegt von Dipl.-Math. Vili Dhamo von der Fakultät II - Mathematik und Naturwissenschaften
More informationMetric Spaces Joseph Muscat 2003 (Last revised May 2009)
1 Distance J Muscat 1 Metric Spaces Joseph Muscat 2003 (Last revised May 2009) (A revised and expanded version of these notes are now published by Springer.) 1 Distance A metric space can be thought of
More informationON SEQUENTIAL CONTINUITY OF COMPOSITION MAPPING. 0. Introduction
ON SEQUENTIAL CONTINUITY OF COMPOSITION MAPPING Abstract. In [1] there was proved a theorem concerning the continuity of the composition mapping, and there was announced a theorem on sequential continuity
More informationProperties of BMO functions whose reciprocals are also BMO
Properties of BMO functions whose reciprocals are also BMO R. L. Johnson and C. J. Neugebauer The main result says that a non-negative BMO-function w, whose reciprocal is also in BMO, belongs to p> A p,and
More informationFUNCTIONAL ANALYSIS LECTURE NOTES: QUOTIENT SPACES
FUNCTIONAL ANALYSIS LECTURE NOTES: QUOTIENT SPACES CHRISTOPHER HEIL 1. Cosets and the Quotient Space Any vector space is an abelian group under the operation of vector addition. So, if you are have studied
More information1. Prove that the empty set is a subset of every set.
1. Prove that the empty set is a subset of every set. Basic Topology Written by Men-Gen Tsai email: b89902089@ntu.edu.tw Proof: For any element x of the empty set, x is also an element of every set since
More informationNONLOCAL PROBLEMS WITH NEUMANN BOUNDARY CONDITIONS
NONLOCAL PROBLEMS WITH NEUMANN BOUNDARY CONDITIONS SERENA DIPIERRO, XAVIER ROS-OTON, AND ENRICO VALDINOCI Abstract. We introduce a new Neumann problem for the fractional Laplacian arising from a simple
More informationWalrasian Demand. u(x) where B(p, w) = {x R n + : p x w}.
Walrasian Demand Econ 2100 Fall 2015 Lecture 5, September 16 Outline 1 Walrasian Demand 2 Properties of Walrasian Demand 3 An Optimization Recipe 4 First and Second Order Conditions Definition Walrasian
More informationStochastic Inventory Control
Chapter 3 Stochastic Inventory Control 1 In this chapter, we consider in much greater details certain dynamic inventory control problems of the type already encountered in section 1.3. In addition to the
More information88 CHAPTER 2. VECTOR FUNCTIONS. . First, we need to compute T (s). a By definition, r (s) T (s) = 1 a sin s a. sin s a, cos s a
88 CHAPTER. VECTOR FUNCTIONS.4 Curvature.4.1 Definitions and Examples The notion of curvature measures how sharply a curve bends. We would expect the curvature to be 0 for a straight line, to be very small
More informationGambling Systems and Multiplication-Invariant Measures
Gambling Systems and Multiplication-Invariant Measures by Jeffrey S. Rosenthal* and Peter O. Schwartz** (May 28, 997.. Introduction. This short paper describes a surprising connection between two previously
More informationFinite dimensional topological vector spaces
Chapter 3 Finite dimensional topological vector spaces 3.1 Finite dimensional Hausdorff t.v.s. Let X be a vector space over the field K of real or complex numbers. We know from linear algebra that the
More informationON COMPLETELY CONTINUOUS INTEGRATION OPERATORS OF A VECTOR MEASURE. 1. Introduction
ON COMPLETELY CONTINUOUS INTEGRATION OPERATORS OF A VECTOR MEASURE J.M. CALABUIG, J. RODRÍGUEZ, AND E.A. SÁNCHEZ-PÉREZ Abstract. Let m be a vector measure taking values in a Banach space X. We prove that
More informationLecture Notes on Measure Theory and Functional Analysis
Lecture Notes on Measure Theory and Functional Analysis P. Cannarsa & T. D Aprile Dipartimento di Matematica Università di Roma Tor Vergata cannarsa@mat.uniroma2.it daprile@mat.uniroma2.it aa 2006/07 Contents
More informationTHE EIGENVALUES OF THE LAPLACIAN ON DOMAINS WITH SMALL SLITS
THE EIGENVALUES OF THE LAPLACIAN ON DOMAINS WITH SMALL SLITS LUC HILLAIRET AND CHRIS JUDGE Abstract. We introduce a small slit into a planar domain and study the resulting effect upon the eigenvalues of
More information1. Introduction. O. MALI, A. MUZALEVSKIY, and D. PAULY
Russ. J. Numer. Anal. Math. Modelling, Vol. 28, No. 6, pp. 577 596 (2013) DOI 10.1515/ rnam-2013-0032 c de Gruyter 2013 Conforming and non-conforming functional a posteriori error estimates for elliptic
More informationLet H and J be as in the above lemma. The result of the lemma shows that the integral
Let and be as in the above lemma. The result of the lemma shows that the integral ( f(x, y)dy) dx is well defined; we denote it by f(x, y)dydx. By symmetry, also the integral ( f(x, y)dx) dy is well defined;
More informationSolutions to Homework 10
Solutions to Homework 1 Section 7., exercise # 1 (b,d): (b) Compute the value of R f dv, where f(x, y) = y/x and R = [1, 3] [, 4]. Solution: Since f is continuous over R, f is integrable over R. Let x
More informationSOLUTIONS TO EXERCISES FOR. MATHEMATICS 205A Part 3. Spaces with special properties
SOLUTIONS TO EXERCISES FOR MATHEMATICS 205A Part 3 Fall 2008 III. Spaces with special properties III.1 : Compact spaces I Problems from Munkres, 26, pp. 170 172 3. Show that a finite union of compact subspaces
More informationIntroduction to Topology
Introduction to Topology Tomoo Matsumura November 30, 2010 Contents 1 Topological spaces 3 1.1 Basis of a Topology......................................... 3 1.2 Comparing Topologies.......................................
More informationAnalytic cohomology groups in top degrees of Zariski open sets in P n
Analytic cohomology groups in top degrees of Zariski open sets in P n Gabriel Chiriacescu, Mihnea Colţoiu, Cezar Joiţa Dedicated to Professor Cabiria Andreian Cazacu on her 80 th birthday 1 Introduction
More informationResearch Article Stability Analysis for Higher-Order Adjacent Derivative in Parametrized Vector Optimization
Hindawi Publishing Corporation Journal of Inequalities and Applications Volume 2010, Article ID 510838, 15 pages doi:10.1155/2010/510838 Research Article Stability Analysis for Higher-Order Adjacent Derivative
More informationand s n (x) f(x) for all x and s.t. s n is measurable if f is. REAL ANALYSIS Measures. A (positive) measure on a measurable space
RAL ANALYSIS A survey of MA 641-643, UAB 1999-2000 M. Griesemer Throughout these notes m denotes Lebesgue measure. 1. Abstract Integration σ-algebras. A σ-algebra in X is a non-empty collection of subsets
More informationUndergraduate Notes in Mathematics. Arkansas Tech University Department of Mathematics
Undergraduate Notes in Mathematics Arkansas Tech University Department of Mathematics An Introductory Single Variable Real Analysis: A Learning Approach through Problem Solving Marcel B. Finan c All Rights
More informationRIGIDITY OF HOLOMORPHIC MAPS BETWEEN FIBER SPACES
RIGIDITY OF HOLOMORPHIC MAPS BETWEEN FIBER SPACES GAUTAM BHARALI AND INDRANIL BISWAS Abstract. In the study of holomorphic maps, the term rigidity refers to certain types of results that give us very specific
More informationMASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 5 9/17/2008 RANDOM VARIABLES
MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 5 9/17/2008 RANDOM VARIABLES Contents 1. Random variables and measurable functions 2. Cumulative distribution functions 3. Discrete
More information1 Norms and Vector Spaces
008.10.07.01 1 Norms and Vector Spaces Suppose we have a complex vector space V. A norm is a function f : V R which satisfies (i) f(x) 0 for all x V (ii) f(x + y) f(x) + f(y) for all x,y V (iii) f(λx)
More information(Basic definitions and properties; Separation theorems; Characterizations) 1.1 Definition, examples, inner description, algebraic properties
Lecture 1 Convex Sets (Basic definitions and properties; Separation theorems; Characterizations) 1.1 Definition, examples, inner description, algebraic properties 1.1.1 A convex set In the school geometry
More informationSumit Chandok and T. D. Narang INVARIANT POINTS OF BEST APPROXIMATION AND BEST SIMULTANEOUS APPROXIMATION
F A S C I C U L I M A T H E M A T I C I Nr 51 2013 Sumit Chandok and T. D. Narang INVARIANT POINTS OF BEST APPROXIMATION AND BEST SIMULTANEOUS APPROXIMATION Abstract. In this paper we generalize and extend
More informationAn explicit inversion formula for the exponetial Radon transform using data from 180
ISSN: 40-567 An explicit inversion formula for the exponetial Radon transform using data from 80 Hans Rullgård Research Reports in Mathematics Number 9, 00 Department of Mathematics Stockholm University
More informationRANDOM INTERVAL HOMEOMORPHISMS. MICHA L MISIUREWICZ Indiana University Purdue University Indianapolis
RANDOM INTERVAL HOMEOMORPHISMS MICHA L MISIUREWICZ Indiana University Purdue University Indianapolis This is a joint work with Lluís Alsedà Motivation: A talk by Yulij Ilyashenko. Two interval maps, applied
More informationLecture L3 - Vectors, Matrices and Coordinate Transformations
S. Widnall 16.07 Dynamics Fall 2009 Lecture notes based on J. Peraire Version 2.0 Lecture L3 - Vectors, Matrices and Coordinate Transformations By using vectors and defining appropriate operations between
More informationNumerical Verification of Optimality Conditions in Optimal Control Problems
Numerical Verification of Optimality Conditions in Optimal Control Problems Dissertation zur Erlangung des naturwissenschaftlichen Doktorgrades der Julius-Maximilians-Universität Würzburg vorgelegt von
More informationt := maxγ ν subject to ν {0,1,2,...} and f(x c +γ ν d) f(x c )+cγ ν f (x c ;d).
1. Line Search Methods Let f : R n R be given and suppose that x c is our current best estimate of a solution to P min x R nf(x). A standard method for improving the estimate x c is to choose a direction
More informationExistence and multiplicity of solutions for a Neumann-type p(x)-laplacian equation with nonsmooth potential. 1 Introduction
Electronic Journal of Qualitative Theory of Differential Equations 20, No. 7, -0; http://www.math.u-szeged.hu/ejqtde/ Existence and multiplicity of solutions for a Neumann-type p(x)-laplacian equation
More informationI. Pointwise convergence
MATH 40 - NOTES Sequences of functions Pointwise and Uniform Convergence Fall 2005 Previously, we have studied sequences of real numbers. Now we discuss the topic of sequences of real valued functions.
More informationEXERCISES PDE 31.10.12-02.11.12. v(x)
EXERCISES PDE 31.1.12-2.11.12 1. Exercise Let U R N 2 be a bounded open set. We say that v C (Ū) is subharmonic iff v in U. (a) Prove that subharmonic functions enjoy the following form of the mean-value
More informationarxiv:math/0501161v1 [math.ds] 11 Jan 2005
ANALYTICITY OF THE SUSCEPTIBILITY FUNCTION FOR UNIMODAL MARKOVIAN MAPS OF THE INTERVAL. by Yunping iang* and David Ruelle**. arxiv:math/0501161v1 [math.ds] 11 an 2005 Abstract. We study the expression
More informationFollow links for Class Use and other Permissions. For more information send email to: permissions@pupress.princeton.edu
COPYRIGHT NOTICE: Ariel Rubinstein: Lecture Notes in Microeconomic Theory is published by Princeton University Press and copyrighted, c 2006, by Princeton University Press. All rights reserved. No part
More information1. Introduction. PROPER HOLOMORPHIC MAPPINGS BETWEEN RIGID POLYNOMIAL DOMAINS IN C n+1
Publ. Mat. 45 (2001), 69 77 PROPER HOLOMORPHIC MAPPINGS BETWEEN RIGID POLYNOMIAL DOMAINS IN C n+1 Bernard Coupet and Nabil Ourimi Abstract We describe the branch locus of proper holomorphic mappings between
More informationTriangle deletion. Ernie Croot. February 3, 2010
Triangle deletion Ernie Croot February 3, 2010 1 Introduction The purpose of this note is to give an intuitive outline of the triangle deletion theorem of Ruzsa and Szemerédi, which says that if G = (V,
More informationLecture notes on Ordinary Differential Equations. Plan of lectures
1 Lecture notes on Ordinary Differential Equations Annual Foundation School, IIT Kanpur, Dec.3-28, 2007. by Dept. of Mathematics, IIT Bombay, Mumbai-76. e-mail: sivaji.ganesh@gmail.com References Plan
More informationAdvanced Microeconomics
Advanced Microeconomics Ordinal preference theory Harald Wiese University of Leipzig Harald Wiese (University of Leipzig) Advanced Microeconomics 1 / 68 Part A. Basic decision and preference theory 1 Decisions
More informationCollege of the Holy Cross, Spring 2009 Math 373, Partial Differential Equations Midterm 1 Practice Questions
College of the Holy Cross, Spring 29 Math 373, Partial Differential Equations Midterm 1 Practice Questions 1. (a) Find a solution of u x + u y + u = xy. Hint: Try a polynomial of degree 2. Solution. Use
More informationAn example of a computable
An example of a computable absolutely normal number Verónica Becher Santiago Figueira Abstract The first example of an absolutely normal number was given by Sierpinski in 96, twenty years before the concept
More informationSome stability results of parameter identification in a jump diffusion model
Some stability results of parameter identification in a jump diffusion model D. Düvelmeyer Technische Universität Chemnitz, Fakultät für Mathematik, 09107 Chemnitz, Germany Abstract In this paper we discuss
More informationInternational Journal of Information Technology, Modeling and Computing (IJITMC) Vol.1, No.3,August 2013
FACTORING CRYPTOSYSTEM MODULI WHEN THE CO-FACTORS DIFFERENCE IS BOUNDED Omar Akchiche 1 and Omar Khadir 2 1,2 Laboratory of Mathematics, Cryptography and Mechanics, Fstm, University of Hassan II Mohammedia-Casablanca,
More informationReference: Introduction to Partial Differential Equations by G. Folland, 1995, Chap. 3.
5 Potential Theory Reference: Introduction to Partial Differential Equations by G. Folland, 995, Chap. 3. 5. Problems of Interest. In what follows, we consider Ω an open, bounded subset of R n with C 2
More informationAn optimal transportation problem with import/export taxes on the boundary
An optimal transportation problem with import/export taxes on the boundary Julián Toledo Workshop International sur les Mathématiques et l Environnement Essaouira, November 2012..................... Joint
More informationMathematics Course 111: Algebra I Part IV: Vector Spaces
Mathematics Course 111: Algebra I Part IV: Vector Spaces D. R. Wilkins Academic Year 1996-7 9 Vector Spaces A vector space over some field K is an algebraic structure consisting of a set V on which are
More informationLecture Notes on Elasticity of Substitution
Lecture Notes on Elasticity of Substitution Ted Bergstrom, UCSB Economics 210A March 3, 2011 Today s featured guest is the elasticity of substitution. Elasticity of a function of a single variable Before
More informationThe Henstock-Kurzweil-Stieltjes type integral for real functions on a fractal subset of the real line
The Henstock-Kurzweil-Stieltjes type integral for real functions on a fractal subset of the real line D. Bongiorno, G. Corrao Dipartimento di Ingegneria lettrica, lettronica e delle Telecomunicazioni,
More informationComplex geodesics in convex tube domains
Complex geodesics in convex tube domains Sylwester Zając Institute of Mathematics, Faculty of Mathematics and Computer Science, Jagiellonian University, Łojasiewicza 6, 30-348 Kraków, Poland sylwester.zajac@im.uj.edu.pl
More informationCHAPTER II THE LIMIT OF A SEQUENCE OF NUMBERS DEFINITION OF THE NUMBER e.
CHAPTER II THE LIMIT OF A SEQUENCE OF NUMBERS DEFINITION OF THE NUMBER e. This chapter contains the beginnings of the most important, and probably the most subtle, notion in mathematical analysis, i.e.,
More informationFINITE DIMENSIONAL ORDERED VECTOR SPACES WITH RIESZ INTERPOLATION AND EFFROS-SHEN S UNIMODULARITY CONJECTURE AARON TIKUISIS
FINITE DIMENSIONAL ORDERED VECTOR SPACES WITH RIESZ INTERPOLATION AND EFFROS-SHEN S UNIMODULARITY CONJECTURE AARON TIKUISIS Abstract. It is shown that, for any field F R, any ordered vector space structure
More informationChapter 2. Parameterized Curves in R 3
Chapter 2. Parameterized Curves in R 3 Def. A smooth curve in R 3 is a smooth map σ : (a, b) R 3. For each t (a, b), σ(t) R 3. As t increases from a to b, σ(t) traces out a curve in R 3. In terms of components,
More informationCHAPTER IV - BROWNIAN MOTION
CHAPTER IV - BROWNIAN MOTION JOSEPH G. CONLON 1. Construction of Brownian Motion There are two ways in which the idea of a Markov chain on a discrete state space can be generalized: (1) The discrete time
More informationON THE NUMBER OF REAL HYPERSURFACES HYPERTANGENT TO A GIVEN REAL SPACE CURVE
Illinois Journal of Mathematics Volume 46, Number 1, Spring 2002, Pages 145 153 S 0019-2082 ON THE NUMBER OF REAL HYPERSURFACES HYPERTANGENT TO A GIVEN REAL SPACE CURVE J. HUISMAN Abstract. Let C be a
More informationMatrix Representations of Linear Transformations and Changes of Coordinates
Matrix Representations of Linear Transformations and Changes of Coordinates 01 Subspaces and Bases 011 Definitions A subspace V of R n is a subset of R n that contains the zero element and is closed under
More information