Alphabeta Math
TheoremStatement: AI-adaptedProof: AI-adaptedPipeline-generated
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Associated Legendre function normalization

Statement

Assume the Axiom of Countable Choice (The Axiom of Countable Choice (ACω)). Let Pl be the Legendre polynomials of Legendre Rodrigues equation, and for l∈N and 0≤m≤l define the function wlm:(−1,1)→R by

wlm(t):=(−1)m(1−t2)m/2 dmdtmPl(t),

where (1−t2)1/2>0 denotes the nonnegative square root. Let σ be the polar surface measure on S2 and let the angle coordinates be Φ(θ,φ)=(sin⁡θcos⁡φ,sin⁡θsin⁡φ,cos⁡θ) (Sphere integrals in angle coordinates, The polar surface set function on the unit sphere). The image of Φ is S2∖C, where C:={(x,0,z)∈S2:x≥0}; C is closed and σ(C)=0 by applying the integral identity of Sphere integrals in angle coordinates to its indicator. For l≥m≥0, define the representative Y~lm:S2→C on S2∖C by Y~lm(Φ(θ,φ)):=blmwlm(cos⁡θ)eimφ2π, and on the cut C by Y~lm(sin⁡θ,0,cos⁡θ):=blmwlm(cos⁡θ)/2π(0≤θ≤π). The limits at φ=0 and 2π agree because m is an integer, and the endpoint values of wlm from (a) give the limits at the poles; hence Y~lm is continuous. For 1≤m≤l set Y~l,−m:=(−1)mY~lm‾, and write Ylm for the class of Y~lm in L2(S2,σ;C). Then:

(a) each wlm extends continuously to [−1,1], with wlm(±1)=0 when m≥1, and on (−1,1) it satisfies the associated Legendre equation −((1−t2)wlm′)′+m21−t2wlm=l(l+1)wlm(−1<t<1), equivalently (1−t2)wlm′′−2twlm′+[l(l+1)−m21−t2]wlm=0;

(b) for fixed m≥0 the functions wlm, l≥m, are pairwise orthogonal in L2((−1,1),dt): ∫−11wlmwkm dt=0 for k≠l; moreover 0<∫−11wlm2 dt<+∞;

(c) with blm:=(∫−11wlm2 dt)−1/2>0 and the representatives just defined, one has ∣Yl,−m∣=∣Ylm∣ and Ylm‾=(−1)mYl,−m for all l and all m with ∣m∣≤l;

(d) the family {Ylm:l≥0, −l≤m≤l} is orthonormal in L2(S2,σ;C) (Real and complex inner product spaces, with the inner product linear in the first argument, The space Lp(μ) as the quotient by null functions): for all (l,m)≠(l′,m′) with ∣m∣≤l, ∣m′∣≤l′, the integral ∫S2YlmYl′m′‾ dσ equals 0, and it equals 1 when (l,m)=(l′,m′);

(e) wl0=Pl, and wll(t)=(−1)l(2l)!2ll!(1−t2)l/2.

The functions Ylm are complex-valued L2-functions on the sphere; no physical interpretation as angular momentum eigenfunctions or spherical harmonics of an atom is asserted.

Facts & Assumptions

Given: the Legendre polynomials Pl; integers l≥0, 0≤m≤l; the functions wlm; the measure σ and the parametrisation Φ; the constants blm and functions Ylm.

[F1]

Pl satisfies (1−t2)Pl′′−2tPl′+l(l+1)Pl=0 on R, is a polynomial of degree l, and Pl(m) is therefore the polynomial function of a degree-(l−m) polynomial (Legendre Rodrigues equation).

[F3]

Sphere integrals in angle coordinates: for Borel F≥0, ∫S2F dσ=∫(0,π)×(0,2π)F(Φ(θ,φ))sin⁡θ d(θ,φ), and the complex L1 version; σ(S2)=4π (Sphere integrals in angle coordinates).

[F4]

Characters of the circle: ∫02πeikφ dφ=2π if k=0 and =0 for every nonzero k∈Z (The trigonometric characters are orthonormal in L2 of the torus, The complex exponential by its power series).

[F5]

A continuous real function on the compact interval [−1,1] is (Riemann and Lebesgue) integrable; products of continuous functions are continuous, and ∫−11g>0 for a continuous g≥0 that is not identically zero (A continuous function on [a,b] is Riemann integrable, by Heine-Cantor and Riemann's criterion, Complex Lp classes and Euclidean test-function conventions).

[F6]

If G:[a,b]→R is differentiable at every point of [a,b] as a function on [a,b] (with one-sided derivatives at a,b) and G′ is integrable on [a,b], then ∫abG′=G(b)−G(a) (The second fundamental theorem: if G is differentiable on [a,b] with G′=f and f is integrable, then ∫abf=G(b)−G(a)).

[F7]

For f,g∈L2 the product fg‾ is in L1 with ∫∣fg‾∣≤∥f∥2∥g∥2; the inner product in L2(S2,σ;C) is linear in the first argument; conjugation satisfies zw‾=z‾ w‾ and ∣z∣2=zz‾ (Cauchy-Schwarz inequality for L2, Real and complex inner product spaces, with the inner product linear in the first argument, Conjugation is an involutive real-field automorphism, zz‾=∣z∣2, and modulus is definite, multiplicative, and subadditive, Complex Lp classes and Euclidean test-function conventions).

[F8]

For the given 0≤m≤l, dmPl/dtm is a polynomial of degree l−m with nonzero leading coefficient, so wlm is not the zero function on (−1,1), because (1−t2)m/2>0 there (Legendre Rodrigues equation, The factorial n! and the falling factorial nk‾, defined by recursion in N).

[F9]

If G:[a,b]→R is continuous on [a,b] and differentiable at every point of (a,b) as a function on [a,b], the mean value theorem gives a c∈(a,b) with G(b)−G(a)=G′(c)(b−a) (The mean value theorem, as the case g(x)=x of Cauchy's: for f continuous on [a,b] with a<b and differentiable on (a,b) there is c∈(a,b) with f(b)−f(a)=f′(c)(b−a)).

Proof

1.1F1F2algebra

[Differentiated equation] Let v:=Pl(m). Differentiating the Legendre equation of [F1] m times with the general Leibniz rule [F2] gives, since the third and higher derivatives of 1−t2 vanish, (1−t2)v′′−2t(m+1)v′+[l(l+1)−m(m+1)]v=0: the coefficient of v′′ is 1−t2; that of v′ collects −2t from differentiating 1−t2 once times the derivative count (m1) and −2t from the term −2tPl′, total −2t(m+1); and that of v collects (m2)(−2)+m(−2)+l(l+1)=−m(m−1)−2m+l(l+1)=l(l+1)−m(m+1).

1.2F1F2algebra

[Continuity at the endpoints] For m=0, wl0=Pl is a polynomial; for m≥1, wlm=(−1)m(1−t2)m/2Q(t) with Q=Pl(m) a polynomial, so ∣wlm(t)∣≤(1−t2)m/2sup⁡[−1,1]∣Q∣→0 as t→±1, and the function extended by wlm(±1):=0 is continuous on [−1,1].

2.1step 1.1F2algebra

[Substitution] Put f(t):=(1−t2)m/2>0 on (−1,1), so that wlm=(−1)mfv; the sign is constant and will be carried along. Put u:=(−1)mv, so w=fu and u satisfies the same homogeneous equation as v in step 1.1. From u=f−1w and [F2], u′=f−1w′−f−2f′w and u′′=f−1w′′−2f−2f′w′+(2f−3(f′)2−f−2f′′)w. Substituting into step 1.1 and multiplying by f gives (1−t2)w′′+[−2f′f(1−t2)−2t(m+1)]w′+[−f′′f(1−t2)+2(f′f)2(1−t2)+2t(m+1)f′f−m(m+1)+l(l+1)]w=0. Direct differentiation gives f′f=−mt1−t2 and therefore f′′f=(f′f)′+(f′f)2=−m−mt2(1−t2)2+m2t2(1−t2)2=m(−1+(m−1)t2)(1−t2)2. Hence the coefficient of w′ is 2mt−2t(m+1)=−2t, and the coefficient of w equals m−m(m+1)t21−t2−m(m+1)+l(l+1)=l(l+1)−m21−t2, using m−m(m+1)t21−t2−m(m+1)=−m21−t2 (expand the left side over the common denominator 1−t2). This is the associated equation.

3.1step 2.1step 1.2F2F5F6F9algebra

[Orthogonality] Fix m and let k≠l, k≥m. Multiply the associated equation for wlm by wkm and that for wkm by wlm, subtract, and use −((1−t2)wlm′)′wkm+((1−t2)wkm′)′wlm=−ddt[(1−t2)(wlm′wkm−wlmwkm′)]: (l(l+1)−k(k+1))wlmwkm=−ddtG(t) with G(t):=(1−t2)(wlm′(t)wkm(t)−wlm(t)wkm′(t)) on (−1,1). By step 1.2, put Ql=Pl(m) and Qk=Pk(m). If m=0, then both w functions are polynomials and the factor 1−t2 gives G(t)→0 at both endpoints. If m≥1, put f(t)=(1−t2)m/2, so f′=−mt(1−t2)m/2−1 and wlm=(−1)mfQl, wkm=(−1)mfQk; since Ql,Qk and their derivatives are bounded on [−1,1], each term in (1−t2)wlm′wkm and (1−t2)wlmwkm′, after the product rule, is bounded by a constant times (1−t2)m or (1−t2)m+1, so again G(t)→0 at both endpoints. Extend G by G(±1)=0; these limits make it continuous on [−1,1]. The subtracted equations give G′(t)=q(t):=−(l(l+1)−k(k+1))wlm(t)wkm(t) on (−1,1), and step 1.2 makes q continuous on [−1,1]. For each x∈(−1,1), [F9] on [x,1] gives (G(1)−G(x))/(1−x)=q(ξx) for some ξx∈(x,1), so as x↑1 this quotient tends to q(1). On [−1,x], [F9] gives (G(x)−G(−1))/(x+1)=q(ηx) for some ηx∈(−1,x); as x↓−1 this quotient tends to q(−1). Thus G is differentiable at every point of [−1,1] with the endpoint convention of [F6], and G′=q is continuous, hence integrable by [F5]. Now [F6] gives ∫−11G′=G(1)−G(−1)=0, so (l(l+1)−k(k+1))∫−11wlmwkm dt=0 and, since l≠k, ∫−11wlmwkm dt=0.

3.2step 2.1step 1.2F5F8algebra

[Positivity and finiteness of the norms] By steps 2.1 and 1.2, wlm2 is continuous on [−1,1], so its integral is finite and nonnegative by [F5]; it is not zero because wlm is not the zero function on (−1,1): Pl(m) is a nonzero polynomial by [F8] and (1−t2)m/2>0 in the interior, so wlm is nonzero at every point where the polynomial does not vanish, and a continuous nonzero function has positive integral by [F5]. Hence 0<∫−11wlm2 dt<+∞ and blm of the Statement is a well-defined positive real.

3.3step 2.1F1F7algebra

[The functions Ylm and their conjugation] The formula in the Statement defines continuous representatives on S2, so each class Ylm belongs to L2(S2,σ) because σ is finite; their chart moduli satisfy ∣Ylm(Φ(θ,φ))∣=blm∣wlm(cos⁡θ)∣/2π. By definition Yl,−m=(−1)mYlm‾, so ∣Yl,−m∣=∣Ylm∣ and, conjugating twice, Yl,−m‾=(−1)mYlm, which is the displayed identity for every m≥0. This establishes (c), and the identities in (e) follow from Pl(0)=Pl and Pl(l)=(2l)!2ll! (the leading coefficient of Pl times l!, by [F1]).

4.1step 3.1step 3.2step 3.3F3F4F7algebra∎

[Orthonormality] Let m,m′≥0. The extensions on C do not affect integrals because σ(C)=0. Thus [F3] applies to the L1 function YlmYl′m′‾, whose integral against σ is the inner product; with [F4], this gives ⟨Ylm,Yl′m′⟩=blmbl′m′12π(∫02πei(m−m′)φdφ)(∫−11wlm(t)wl′m′(t) dt)=δmm′ blmbl′m∫−11wlmwl′m dt, which by steps 3.1 and 3.2 equals 0 for l≠l′ and 1 for l=l′. For entries with negative lower index, write the inner product as a product of conjugation operations using [F7]: ⟨Yl,−m,Yl′,−m′⟩=(−1)m+m′⟨Ylm‾,Yl′m′‾⟩=(−1)m+m′⟨Ylm,Yl′m′⟩‾, and ⟨Ylm,Yl′,−m′⟩=(−1)m′∫S2YlmYl′m′ dσ, whose rectangle form has φ-factor ∫02πei(m+m′)φdφ, vanishing unless m=m′=0 by [F4], in which case it reduces to the already treated case; collecting the cases gives orthonormality for all pairs. This proves (d), and (a), (b) were proved in steps 2.1, 3.1 and 3.2.

Depends on

Dependency tree · 2 levels

Sources