Contents · Previous · Next

ZR=−cR(10+x⋅∇log⁡w)−2cR′(s−s0). Z_R=-c_R(10+x\cdot\nabla\log w)-2c_R'(s-s_0).

Here cR′≤0c_R'\le0, so ZR≥5cRZ_R\ge5c_R beyond one fixed radius. Its compact negative part is controlled by NhN_h. The integration by parts is legitimate before physical phase integrability is known: a further cutoff at radius TT has error bounded by CRT−2∫T<∣x∣<2Twχ2/(1+∣x∣2)dx→0C_RT^{-2}\int_{T<|x|<2T}w\chi^2/(1+|x|^2)dx\to0. There is no boundary at the smooth plateau join.

Uniformly in R≥1R\ge1, direct differentiation gives

∣∇FR∣2FR≤C,∣∇2FR∣2FR≤C/s,∣∇⋅(w∇FR)∣w≤C,∣∇FR∣2cR≤Cs,∥V∥∣∇FR∣≤CcR/s. \frac{|\nabla F_R|^2}{F_R}\le C,\quad \frac{|\nabla^2F_R|^2}{F_R}\le C/s,\quad \frac{|\nabla\cdot(w\nabla F_R)|}{w}\le C,\quad \frac{|\nabla F_R|^2}{c_R}\le Cs,\quad \|V\||\nabla F_R|\le Cc_R/\sqrt s.

Integrating E1E_1 by parts bounds it by CNhCN_h. Young's inequality and (N11) give ∣E2∣≤δ∫wFR∣∇X∣2dx+CδNh|E_2|\le\delta\int wF_R|\nabla X|^2dx+C_\delta N_h and ∣E3∣≤CNh|E_3|\le CN_h. The tested source is at most

δ∫wFR∣X∣2dx+δ∫wcRχ2dx+Cδ∫ws∣f∣2dx. \delta\int wF_R|X|^2dx+\delta\int wc_R\chi^2dx+ C_\delta\int ws|f|^2dx.

Choose δ\delta smaller than the positive leading energy coefficients and the exterior phase coefficient. Absorption yields, independently of RR,

∫wFR(∣X∣2+∣∇X∣2+∣∇χ∣2)dx+∫wcRχ2dx≤C(Nh+∫ws∣f∣2dx). \int wF_R(|X|^2+|\nabla X|^2+|\nabla\chi|^2)dx+ \int wc_R\chi^2dx\le C\left(N_h+\int ws|f|^2dx\right).

As R→∞R\to\infty, FR↑sF_R\uparrow s and cR↑1c_R\uparrow1, proving (N10). The additional source moment and absence of a phase source are essential. This is not a bounded inverse for arbitrary physical L2L^2 forcing, and it is consistent with the unchanged essential zero threshold.

Physical Domain, Chain Counts And Inertia

Estimate (N10), multiplication by w\sqrt w, and the bounded logarithmic derivative of ww place the solution in the physical H1H^1 form domain. The normalized affine forcing lies in L2L^2. The weak equation therefore places the solution in D(A)=H2\mathcal D(\mathscr A)=H^2. Equivalently, the normalized principal Laplacians have positive constant coefficients and all remaining first-order and zeroth-order coefficients are bounded; the equation puts their Laplacians in L2L^2, giving H2H^2 by the full-space Fourier characterization. Thus the result solves the physical operator equation, rather than only a weighted formal divergence equation. The reduced solution is unique modulo constant common phase; testing the difference with itself sets X=0X=0 and ∇χ=0\nabla\chi=0. Complete correctors are unique modulo the full static kernel.

Every kernel direction with zero common-phase coefficient now has a partner, after subtracting (N6). The independent support obstruction excludes every direction with nonzero common-phase coefficient. The exact classification is

Branch Static Dimension Length-Two Chains Generalized Dimension

Contents · Previous · Next